Datasets:

sswt

/

arxiv_sample_50K

like 0

To Point or Not to Point Understanding How Abstractive Summarizers Paraphrase Text ; Abstractive neural summarization models have seen great improvements in recent years, as shown by ROUGE scores of the generated summaries. But despite these improved metrics, there is limited understanding of the strategies different models employ, and how those strategies relate their understanding of language. To understand this better, we run several experiments to characterize how one popular abstractive model, the pointergenerator model of See et al. 2017, uses its explicit copygeneration switch to control its level of abstraction generation vs extraction copying. On an extractivebiased dataset, the model utilizes syntactic boundaries to truncate sentences that are otherwise often copied verbatim. When we modify the copygeneration switch and force the model to generate, only simple paraphrasing abilities are revealed alongside factual inaccuracies and hallucinations. On an abstractivebiased dataset, the model copies infrequently but shows similarly limited abstractive abilities. In line with previous research, these results suggest that abstractive summarization models lack the semantic understanding necessary to generate paraphrases that are both abstractive and faithful to the source document.

Network Generation with Differential Privacy ; We consider the problem of generating private synthetic versions of realworld graphs containing private information while maintaining the utility of generated graphs. Differential privacy is a gold standard for data privacy, and the introduction of the differentially private stochastic gradient descent DPSGD algorithm has facilitated the training of private neural models in a number of domains. Recent advances in graph generation via deep generative networks have produced several high performing models. We evaluate and compare stateoftheart models including adjacency matrix based models and edge based models, and show a practical implementation that favours the edgelist approach utilizing the Gaussian noise mechanism when evaluated on commonly used graph datasets. Based on our findings, we propose a generative model that can reproduce the properties of realworld networks while maintaining edgedifferential privacy. The proposed model is based on a stochastic neural network that generates discrete edgelist samples and is trained using the Wasserstein GAN objective with the DPSGD optimizer. Being the first approach to combine these beneficial properties, our model contributes to further research on graph data privacy.

Riemannian ScoreBased Generative Modelling ; Scorebased generative models SGMs are a powerful class of generative models that exhibit remarkable empirical performance. Scorebased generative modelling SGM consists of a noising'' stage, whereby a diffusion is used to gradually add Gaussian noise to data, and a generative model, which entails a denoising'' process defined by approximating the timereversal of the diffusion. Existing SGMs assume that data is supported on a Euclidean space, i.e. a manifold with flat geometry. In many domains such as robotics, geoscience or protein modelling, data is often naturally described by distributions living on Riemannian manifolds and current SGM techniques are not appropriate. We introduce here Riemannian Scorebased Generative Models RSGMs, a class of generative models extending SGMs to Riemannian manifolds. We demonstrate our approach on a variety of manifolds, and in particular with earth and climate science spherical data.

Can Pushforward Generative Models Fit Multimodal Distributions ; Many generative models synthesize data by transforming a standard Gaussian random variable using a deterministic neural network. Among these models are the Variational Autoencoders and the Generative Adversarial Networks. In this work, we call them pushforward models and study their expressivity. We show that the Lipschitz constant of these generative networks has to be large in order to fit multimodal distributions. More precisely, we show that the total variation distance and the KullbackLeibler divergence between the generated and the data distribution are bounded from below by a constant depending on the mode separation and the Lipschitz constant. Since constraining the Lipschitz constants of neural networks is a common way to stabilize generative models, there is a provable tradeoff between the ability of pushforward models to approximate multimodal distributions and the stability of their training. We validate our findings on onedimensional and image datasets and empirically show that generative models consisting of stacked networks with stochastic input at each step, such as diffusion models do not suffer of such limitations.

Your ViT is Secretly a Hybrid DiscriminativeGenerative Diffusion Model ; Diffusion Denoising Probability Models DDPM and Vision Transformer ViT have demonstrated significant progress in generative tasks and discriminative tasks, respectively, and thus far these models have largely been developed in their own domains. In this paper, we establish a direct connection between DDPM and ViT by integrating the ViT architecture into DDPM, and introduce a new generative model called Generative ViT GenViT. The modeling flexibility of ViT enables us to further extend GenViT to hybrid discriminativegenerative modeling, and introduce a Hybrid ViT HybViT. Our work is among the first to explore a single ViT for image generation and classification jointly. We conduct a series of experiments to analyze the performance of proposed models and demonstrate their superiority over prior stateofthearts in both generative and discriminative tasks. Our code and pretrained models can be found in httpsgithub.comsndnyangDiffusionViT .

How good are deep models in understanding the generated images ; My goal in this paper is twofold to study how well deep models can understand the images generated by DALLE 2 and Midjourney, and to quantitatively evaluate these generative models. Two sets of generated images are collected for object recognition and visual question answering VQA tasks. On object recognition, the best model, out of 10 stateoftheart object recognition models, achieves about 60 and 80 top1 and top5 accuracy, respectively. These numbers are much lower than the best accuracy on the ImageNet dataset 91 and 99. On VQA, the OFA model scores 77.3 on answering 241 binary questions across 50 images. This model scores 94.7 on the binary VQAv2 dataset. Humans are able to recognize the generated images and answer questions on them easily. We conclude that a deep models struggle to understand the generated content, and may do better after finetuning, and b there is a large distribution shift between the generated images and the real photographs. The distribution shift appears to be categorydependent. Data is available at httpsdrive.google.comfiled1n2nCiaXtYJRRF2R73LNE3zggeUHeH0viewuspsharing.

CHeart A Conditional SpatioTemporal Generative Model for Cardiac Anatomy ; Two key questions in cardiac image analysis are to assess the anatomy and motion of the heart from images; and to understand how they are associated with nonimaging clinical factors such as gender, age and diseases. While the first question can often be addressed by image segmentation and motion tracking algorithms, our capability to model and to answer the second question is still limited. In this work, we propose a novel conditional generative model to describe the 4D spatiotemporal anatomy of the heart and its interaction with nonimaging clinical factors. The clinical factors are integrated as the conditions of the generative modelling, which allows us to investigate how these factors influence the cardiac anatomy. We evaluate the model performance in mainly two tasks, anatomical sequence completion and sequence generation. The model achieves a high performance in anatomical sequence completion, comparable to or outperforming other stateoftheart generative models. In terms of sequence generation, given clinical conditions, the model can generate realistic synthetic 4D sequential anatomies that share similar distributions with the real data.

SlotDiffusion ObjectCentric Generative Modeling with Diffusion Models ; Objectcentric learning aims to represent visual data with a set of object entities a.k.a. slots, providing structured representations that enable systematic generalization. Leveraging advanced architectures like Transformers, recent approaches have made significant progress in unsupervised object discovery. In addition, slotbased representations hold great potential for generative modeling, such as controllable image generation and object manipulation in image editing. However, current slotbased methods often produce blurry images and distorted objects, exhibiting poor generative modeling capabilities. In this paper, we focus on improving slottoimage decoding, a crucial aspect for highquality visual generation. We introduce SlotDiffusion an objectcentric Latent Diffusion Model LDM designed for both image and video data. Thanks to the powerful modeling capacity of LDMs, SlotDiffusion surpasses previous slot models in unsupervised object segmentation and visual generation across six datasets. Furthermore, our learned object features can be utilized by existing objectcentric dynamics models, improving video prediction quality and downstream temporal reasoning tasks. Finally, we demonstrate the scalability of SlotDiffusion to unconstrained realworld datasets such as PASCAL VOC and COCO, when integrated with selfsupervised pretrained image encoders.

GPTFL Generative Pretrained ModelAssisted Federated Learning ; In this work, we propose GPTFL, a generative pretrained modelassisted federated learning FL framework. At its core, GPTFL leverages generative pretrained models to generate diversified synthetic data. These generated data are used to train a downstream model on the server, which is then finetuned with private client data under the standard FL framework. We show that GPTFL consistently outperforms stateoftheart FL methods in terms of model test accuracy, communication efficiency, and client sampling efficiency. Through comprehensive ablation analysis, we discover that the downstream model generated by synthetic data plays a crucial role in controlling the direction of gradient diversity during FL training, which enhances convergence speed and contributes to the notable accuracy boost observed with GPTFL. Also, regardless of whether the target data falls within or outside the domain of the pretrained generative model, GPTFL consistently achieves significant performance gains, surpassing the results obtained by models trained solely with FL or synthetic data.

Exploring an LM to generate Prolog Predicates from Mathematics Questions ; Recently, there has been a surge in interest in NLP driven by ChatGPT. ChatGPT, a transformerbased generative language model of substantial scale, exhibits versatility in performing various tasks based on natural language. Nevertheless, large language models often exhibit poor performance in solving mathematics questions that require reasoning. Prior research has demonstrated the effectiveness of chainofthought prompting in enhancing reasoning capabilities. Now, we aim to investigate whether finetuning a model for the generation of Prolog codes, a logic language, and subsequently passing these codes to a compiler can further improve accuracy. Consequently, we employ chainofthought to finetune LLaMA7B as a baseline model and develop other finetuned LLaMA7B models for the generation of Prolog code, Prolog code chainofthought, and chainofthought Prolog code, respectively. The results reveal that the Prolog generation model surpasses the baseline in performance, while the combination generation models do not yield significant improvements. The Prolog corpus based on GSM8K and the correspondingly finetuned Prolog generation model based on LLaMA7B are released to the research community.

Perceptual Generative Autoencoders ; Modern generative models are usually designed to match target distributions directly in the data space, where the intrinsic dimension of data can be much lower than the ambient dimension. We argue that this discrepancy may contribute to the difficulties in training generative models. We therefore propose to map both the generated and target distributions to a latent space using the encoder of a standard autoencoder, and train the generator or decoder to match the target distribution in the latent space. Specifically, we enforce the consistency in both the data space and the latent space with theoretically justified data and latent reconstruction losses. The resulting generative model, which we call a perceptual generative autoencoder PGA, is then trained with a maximum likelihood or variational autoencoder VAE objective. With maximum likelihood, PGAs generalize the idea of reversible generative models to unrestricted neural network architectures and arbitrary number of latent dimensions. When combined with VAEs, PGAs substantially improve over the baseline VAEs in terms of sample quality. Compared to other autoencoderbased generative models using simple priors, PGAs achieve stateoftheart FID scores on CIFAR10 and CelebA.

What is the Reward for Handwriting Handwriting Generation by Imitation Learning ; Analyzing the handwriting generation process is an important issue and has been tackled by various generation models, such as kinematics based models and stochastic models. In this study, we use a reinforcement learning RL framework to realize handwriting generation with the careful future planning ability. In fact, the handwriting process of human beings is also supported by their future planning ability; for example, the ability is necessary to generate a closed trajectory like '0' because any shortsighted model, such as a Markovian model, cannot generate it. For the algorithm, we employ generative adversarial imitation learning GAIL. Typical RL algorithms require the manual definition of the reward function, which is very crucial to control the generation process. In contrast, GAIL trains the reward function along with the other modules of the framework. In other words, through GAIL, we can understand the reward of the handwriting generation process from handwriting examples. Our experimental results qualitatively and quantitatively show that the learned reward catches the trends in handwriting generation and thus GAIL is well suited for the acquisition of handwriting behavior.

Neural RuleExecution Tracking Machine For TransformerBased Text Generation ; SequencetoSequence S2S neural text generation models, especially the pretrained ones e.g., BART and T5, have exhibited compelling performance on various natural language generation tasks. However, the blackbox nature of these models limits their application in tasks where specific rules e.g., controllable constraints, prior knowledge need to be executed. Previous works either design specific model structure e.g., Copy Mechanism corresponding to the rule the generated output should include certain words in the source input or implement specialized inference algorithm e.g., Constrained Beam Search to execute particular rules through the text generation. These methods require careful design casebycase and are difficult to support multiple rules concurrently. In this paper, we propose a novel module named Neural RuleExecution Tracking Machine that can be equipped into various transformerbased generators to leverage multiple rules simultaneously to guide the neural generation model for superior generation performance in a unified and scalable way. Extensive experimental results on several benchmarks verify the effectiveness of our proposed model in both controllable and general text generation.

HighFidelity Synthesis with Disentangled Representation ; Learning disentangled representation of data without supervision is an important step towards improving the interpretability of generative models. Despite recent advances in disentangled representation learning, existing approaches often suffer from the tradeoff between representation learning and generation performance i.e. improving generation quality sacrifices disentanglement performance. We propose an InformationDistillation Generative Adversarial Network IDGAN, a simple yet generic framework that easily incorporates the existing stateoftheart models for both disentanglement learning and highfidelity synthesis. Our method learns disentangled representation using VAEbased models, and distills the learned representation with an additional nuisance variable to the separate GANbased generator for highfidelity synthesis. To ensure that both generative models are aligned to render the same generative factors, we further constrain the GAN generator to maximize the mutual information between the learned latent code and the output. Despite the simplicity, we show that the proposed method is highly effective, achieving comparable image generation quality to the stateoftheart methods using the disentangled representation. We also show that the proposed decomposition leads to an efficient and stable model design, and we demonstrate photorealistic highresolution image synthesis results 1024x1024 pixels for the first time using the disentangled representations.

Viable Threat on News Reading Generating Biased News Using Natural Language Models ; Recent advancements in natural language generation has raised serious concerns. Highperformance language models are widely used for language generation tasks because they are able to produce fluent and meaningful sentences. These models are already being used to create fake news. They can also be exploited to generate biased news, which can then be used to attack news aggregators to change their reader's behavior and influence their bias. In this paper, we use a threat model to demonstrate that the publicly available language models can reliably generate biased news content based on an input original news. We also show that a large number of highquality biased news articles can be generated using controllable text generation. A subjective evaluation with 80 participants demonstrated that the generated biased news is generally fluent, and a bias evaluation with 24 participants demonstrated that the bias left or right is usually evident in the generated articles and can be easily identified.

DYPLOC Dynamic Planning of Content Using Mixed Language Models for Text Generation ; We study the task of longform opinion text generation, which faces at least two distinct challenges. First, existing neural generation models fall short of coherence, thus requiring efficient content planning. Second, diverse types of information are needed to guide the generator to cover both subjective and objective content. To this end, we propose DYPLOC, a generation framework that conducts dynamic planning of content while generating the output based on a novel design of mixed language models. To enrich the generation with diverse content, we further propose to use large pretrained models to predict relevant concepts and to generate claims. We experiment with two challenging tasks on newly collected datasets 1 argument generation with Reddit ChangeMyView, and 2 writing articles using New York Times' Opinion section. Automatic evaluation shows that our model significantly outperforms competitive comparisons. Human judges further confirm that our generations are more coherent with richer content.

COINS Dynamically Generating COntextualized Inference Rules for Narrative Story Completion ; Despite recent successes of large pretrained language models in solving reasoning tasks, their inference capabilities remain opaque. We posit that such models can be made more interpretable by explicitly generating interim inference rules, and using them to guide the generation of taskspecific textual outputs. In this paper we present COINS, a recursive inference framework that i iteratively reads context sentences, ii dynamically generates contextualized inference rules, encodes them, and iii uses them to guide taskspecific output generation. We apply COINS to a Narrative Story Completion task that asks a model to complete a story with missing sentences, to produce a coherent story with plausible logical connections, causal relationships, and temporal dependencies. By modularizing inference and sentence generation steps in a recurrent model, we aim to make reasoning steps and their effects on next sentence generation transparent. Our automatic and manual evaluations show that the model generates better story sentences than SOTA baselines, especially in terms of coherence. We further demonstrate improved performance over strong pretrained LMs in generating commonsense inference rules. The recursive nature of COINS holds the potential for controlled generation of longer sequences.

CRASH Raw Audio Scorebased Generative Modeling for Controllable Highresolution Drum Sound Synthesis ; In this paper, we propose a novel scorebase generative model for unconditional raw audio synthesis. Our proposal builds upon the latest developments on diffusion process modeling with stochastic differential equations, which already demonstrated promising results on image generation. We motivate novel heuristics for the choice of the diffusion processes better suited for audio generation, and consider the use of a conditional UNet to approximate the score function. While previous approaches on diffusion models on audio were mainly designed as speech vocoders in medium resolution, our method termed CRASH Controllable Raw Audio Synthesis with Highresolution allows us to generate short percussive sounds in 44.1kHz in a controllable way. Through extensive experiments, we showcase on a drum sound generation task the numerous sampling schemes offered by our method unconditional generation, deterministic generation, inpainting, interpolation, variations, classconditional sampling and propose the classmixing sampling, a novel way to generate hybrid sounds. Our proposed method closes the gap with GANbased methods on raw audio, while offering more flexible generation capabilities with lighter and easiertotrain models.

A Kernelised Stein Statistic for Assessing Implicit Generative Models ; Synthetic data generation has become a key ingredient for training machine learning procedures, addressing tasks such as data augmentation, analysing privacysensitive data, or visualising representative samples. Assessing the quality of such synthetic data generators hence has to be addressed. As deep generative models for synthetic data often do not admit explicit probability distributions, classical statistical procedures for assessing model goodnessoffit may not be applicable. In this paper, we propose a principled procedure to assess the quality of a synthetic data generator. The procedure is a kernelised Stein discrepancy KSDtype test which is based on a nonparametric Stein operator for the synthetic data generator of interest. This operator is estimated from samples which are obtained from the synthetic data generator and hence can be applied even when the model is only implicit. In contrast to classical testing, the sample size from the synthetic data generator can be as large as desired, while the size of the observed data, which the generator aims to emulate is fixed. Experimental results on synthetic distributions and trained generative models on synthetic and real datasets illustrate that the method shows improved power performance compared to existing approaches.

Articulation GAN Unsupervised modeling of articulatory learning ; Generative deep neural networks are widely used for speech synthesis, but most existing models directly generate waveforms or spectral outputs. Humans, however, produce speech by controlling articulators, which results in the production of speech sounds through physical properties of sound propagation. We introduce the Articulatory Generator to the Generative Adversarial Network paradigm, a new unsupervised generative model of speech productionsynthesis. The Articulatory Generator more closely mimics human speech production by learning to generate articulatory representations electromagnetic articulography or EMA in a fully unsupervised manner. A separate pretrained physical model ema2wav then transforms the generated EMA representations to speech waveforms, which get sent to the Discriminator for evaluation. Articulatory analysis suggests that the network learns to control articulators in a similar manner to humans during speech production. Acoustic analysis of the outputs suggests that the network learns to generate words that are both present and absent in the training distribution. We additionally discuss implications of articulatory representations for cognitive models of human language and speech technology in general.

SILVR Guided Diffusion for Molecule Generation ; Computationally generating novel synthetically accessible compounds with high affinity and low toxicity is a great challenge in drug design. Machinelearning models beyond conventional pharmacophoric methods have shown promise in generating novel small molecule compounds, but require significant tuning for a specific protein target. Here, we introduce a method called selective iterative latent variable refinement SILVR for conditioning an existing diffusionbased equivariant generative model without retraining. The model allows the generation of new molecules that fit into a binding site of a protein based on fragment hits. We use the SARSCoV2 Main protease fragments from Diamond XChem that form part of the COVID Moonshot project as a reference dataset for conditioning the molecule generation. The SILVR rate controls the extent of conditioning and we show that moderate SILVR rates make it possible to generate new molecules of similar shape to the original fragments, meaning that the new molecules fit the binding site without knowledge of the protein. We can also merge up to 3 fragments into a new molecule without affecting the quality of molecules generated by the underlying generative model. Our method is generalizable to any protein target with known fragments and any diffusionbased model for molecule generation.

You Can Generate It Again Datatotext Generation with Verification and Correction Prompting ; Despite significant advancements in existing models, generating text descriptions from structured data input, known as datatotext generation, remains a challenging task. In this paper, we propose a novel approach that goes beyond traditional oneshot generation methods by introducing a multistep process consisting of generation, verification, and correction stages. Our approach, VCPVerification and Correction Prompting, begins with the model generating an initial output. We then proceed to verify the correctness of different aspects of the generated text. The observations from the verification step are converted into a specialized errorindication prompt, which instructs the model to regenerate the output while considering the identified errors. To enhance the model's correction ability, we have developed a carefully designed training procedure. This procedure enables the model to incorporate feedback from the errorindication prompt, resulting in improved output generation. Through experimental results, we demonstrate that our approach effectively reduces slot error rates while maintaining the overall quality of the generated text.

Personaaware Generative Model for Codemixed Language ; Codemixing and scriptmixing are prevalent across online social networks and multilingual societies. However, a user's preference toward codemixing depends on the socioeconomic status, demographics of the user, and the local context, which existing generative models mostly ignore while generating codemixed texts. In this work, we make a pioneering attempt to develop a personaaware generative model to generate texts resembling reallife codemixed texts of individuals. We propose a Personaaware Generative Model for Codemixed Generation, PARADOX, a novel Transformerbased encoderdecoder model that encodes an utterance conditioned on a user's persona and generates codemixed texts without monolingual reference data. We propose an alignment module that recalibrates the generated sequence to resemble reallife codemixed texts. PARADOX generates codemixed texts that are semantically more meaningful and linguistically more valid. To evaluate the personification capabilities of PARADOX, we propose four new metrics CM BLEU, CM Rouge1, CM RougeL and CM KS. On average, PARADOX achieves 1.6 points better CM BLEU, 47 better perplexity and 32 better semantic coherence than the nonpersonabased counterparts.

IRGAN A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models ; This paper provides a unified account of two schools of thinking in information retrieval modelling the generative retrieval focusing on predicting relevant documents given a query, and the discriminative retrieval focusing on predicting relevancy given a querydocument pair. We propose a game theoretical minimax game to iteratively optimise both models. On one hand, the discriminative model, aiming to mine signals from labelled and unlabelled data, provides guidance to train the generative model towards fitting the underlying relevance distribution over documents given the query. On the other hand, the generative model, acting as an attacker to the current discriminative model, generates difficult examples for the discriminative model in an adversarial way by minimising its discrimination objective. With the competition between these two models, we show that the unified framework takes advantage of both schools of thinking i the generative model learns to fit the relevance distribution over documents via the signals from the discriminative model, and ii the discriminative model is able to exploit the unlabelled data selected by the generative model to achieve a better estimation for document ranking. Our experimental results have demonstrated significant performance gains as much as 23.96 on Precision5 and 15.50 on MAP over strong baselines in a variety of applications including web search, item recommendation, and question answering.

Robustness Analysis of Deep Learning Models for Population Synthesis ; Deep generative models have become useful for synthetic data generation, particularly population synthesis. The models implicitly learn the probability distribution of a dataset and can draw samples from a distribution. Several models have been proposed, but their performance is only tested on a single crosssectional sample. The implementation of population synthesis on single datasets is seen as a drawback that needs further studies to explore the robustness of the models on multiple datasets. While comparing with the real data can increase trust and interpretability of the models, techniques to evaluate deep generative models' robustness for population synthesis remain underexplored. In this study, we present bootstrap confidence interval for the deep generative models, an approach that computes efficient confidence intervals for mean errors predictions to evaluate the robustness of the models to multiple datasets. Specifically, we adopt the tabularbased Composite Travel Generative Adversarial Network CTGAN and Variational Autoencoder VAE, to estimate the distribution of the population, by generating agents that have tabular data using several samples over time from the same study area. The models are implemented on multiple travel diaries of Montreal Origin Destination Survey of 2008, 2013, and 2018 and compare the predictive performance under varying sample sizes from multiple surveys. Results show that the predictive errors of CTGAN have narrower confidence intervals indicating its robustness to multiple datasets of the varying sample sizes when compared to VAE. Again, the evaluation of model robustness against varying sample size shows a minimal decrease in model performance with decrease in sample size. This study directly supports agentbased modelling by enabling finer synthetic generation of populations in a reliable environment.

DiffInstruct A Universal Approach for Transferring Knowledge From Pretrained Diffusion Models ; Due to the ease of training, ability to scale, and high sample quality, diffusion models DMs have become the preferred option for generative modeling, with numerous pretrained models available for a wide variety of datasets. Containing intricate information about data distributions, pretrained DMs are valuable assets for downstream applications. In this work, we consider learning from pretrained DMs and transferring their knowledge to other generative models in a datafree fashion. Specifically, we propose a general framework called DiffInstruct to instruct the training of arbitrary generative models as long as the generated samples are differentiable with respect to the model parameters. Our proposed DiffInstruct is built on a rigorous mathematical foundation where the instruction process directly corresponds to minimizing a novel divergence we call Integral KullbackLeibler IKL divergence. IKL is tailored for DMs by calculating the integral of the KL divergence along a diffusion process, which we show to be more robust in comparing distributions with misaligned supports. We also reveal nontrivial connections of our method to existing works such as DreamFusion, and generative adversarial training. To demonstrate the effectiveness and universality of DiffInstruct, we consider two scenarios distilling pretrained diffusion models and refining existing GAN models. The experiments on distilling pretrained diffusion models show that DiffInstruct results in stateoftheart singlestep diffusionbased models. The experiments on refining GAN models show that the DiffInstruct can consistently improve the pretrained generators of GAN models across various settings.

A Test of Relative Similarity For Model Selection in Generative Models ; Probabilistic generative models provide a powerful framework for representing data that avoids the expense of manual annotation typically needed by discriminative approaches. Model selection in this generative setting can be challenging, however, particularly when likelihoods are not easily accessible. To address this issue, we introduce a statistical test of relative similarity, which is used to determine which of two models generates samples that are significantly closer to a realworld reference dataset of interest. We use as our test statistic the difference in maximum mean discrepancies MMDs between the reference dataset and each model dataset, and derive a powerful, lowvariance test based on the joint asymptotic distribution of the MMDs between each referencemodel pair. In experiments on deep generative models, including the variational autoencoder and generative moment matching network, the tests provide a meaningful ranking of model performance as a function of parameter and training settings.

Modelbased Adversarial Imitation Learning ; Generative adversarial learning is a popular new approach to training generative models which has been proven successful for other related problems as well. The general idea is to maintain an oracle D that discriminates between the expert's data distribution and that of the generative model G. The generative model is trained to capture the expert's distribution by maximizing the probability of D misclassifying the data it generates. Overall, the system is emphdifferentiable endtoend and is trained using basic backpropagation. This type of learning was successfully applied to the problem of policy imitation in a modelfree setup. However, a modelfree approach does not allow the system to be differentiable, which requires the use of highvariance gradient estimations. In this paper we introduce the Model based Adversarial Imitation Learning MAIL algorithm. A modelbased approach for the problem of adversarial imitation learning. We show how to use a forward model to make the system fully differentiable, which enables us to train policies using the stochastic gradient of D. Moreover, our approach requires relatively few environment interactions, and fewer hyperparameters to tune. We test our method on the MuJoCo physics simulator and report initial results that surpass the current stateoftheart.

Defending Neural Backdoors via Generative Distribution Modeling ; Neural backdoor attack is emerging as a severe security threat to deep learning, while the capability of existing defense methods is limited, especially for complex backdoor triggers. In the work, we explore the space formed by the pixel values of all possible backdoor triggers. An original trigger used by an attacker to build the backdoored model represents only a point in the space. It then will be generalized into a distribution of valid triggers, all of which can influence the backdoored model. Thus, previous methods that model only one point of the trigger distribution is not sufficient. Getting the entire trigger distribution, e.g., via generative modeling, is a key to effective defense. However, existing generative modeling techniques for image generation are not applicable to the backdoor scenario as the trigger distribution is completely unknown. In this work, we propose maxentropy staircase approximator MESA, an algorithm for highdimensional samplingfree generative modeling and use it to recover the trigger distribution. We also develop a defense technique to remove the triggers from the backdoored model. Our experiments on Cifar10100 dataset demonstrate the effectiveness of MESA in modeling the trigger distribution and the robustness of the proposed defense method.

Learning Latent Space EnergyBased Prior Model for Molecule Generation ; Deep generative models have recently been applied to molecule design. If the molecules are encoded in linear SMILES strings, modeling becomes convenient. However, models relying on string representations tend to generate invalid samples and duplicates. Prior work addressed these issues by building models on chemicallyvalid fragments or explicitly enforcing chemical rules in the generation process. We argue that an expressive model is sufficient to implicitly and automatically learn the complicated chemical rules from the data, even if molecules are encoded in simple characterlevel SMILES strings. We propose to learn latent space energybased prior model with SMILES representation for molecule modeling. Our experiments show that our method is able to generate molecules with validity and uniqueness competitive with stateoftheart models. Interestingly, generated molecules have structural and chemical features whose distributions almost perfectly match those of the real molecules.

The Power of Fragmentation A Hierarchical Transformer Model for Structural Segmentation in Symbolic Music Generation ; Symbolic Music Generation relies on the contextual representation capabilities of the generative model, where the most prevalent approach is the Transformerbased model. The learning of musical context is also related to the structural elements in music, i.e. intro, verse, and chorus, which are currently overlooked by the research community. In this paper, we propose a hierarchical Transformer model to learn multiscale contexts in music. In the encoding phase, we first designed a Fragment Scope Localization layer to syncopate the music into chords and sections. Then, we use a multiscale attention mechanism to learn note, chord, and sectionlevel contexts. In the decoding phase, we proposed a hierarchical Transformer model that uses finedecoders to generate sections in parallel and a coarsedecoder to decode the combined music. We also designed a Music Style Normalization layer to achieve a consistent music style between the generated sections. Our model is evaluated on two open MIDI datasets, and experiments show that our model outperforms the best contemporary music generative models. More excitingly, the visual evaluation shows that our model is superior in melody reuse, resulting in more realistic music.

On the Strong Correlation Between Model Invariance and Generalization ; Generalization and invariance are two essential properties of any machine learning model. Generalization captures a model's ability to classify unseen data while invariance measures consistency of model predictions on transformations of the data. Existing research suggests a positive relationship a model generalizing well should be invariant to certain visual factors. Building on this qualitative implication we make two contributions. First, we introduce effective invariance EI, a simple and reasonable measure of model invariance which does not rely on image labels. Given predictions on a test image and its transformed version, EI measures how well the predictions agree and with what level of confidence. Second, using invariance scores computed by EI, we perform largescale quantitative correlation studies between generalization and invariance, focusing on rotation and grayscale transformations. From a modelcentric view, we observe generalization and invariance of different models exhibit a strong linear relationship, on both indistribution and outofdistribution datasets. From a datasetcentric view, we find a certain model's accuracy and invariance linearly correlated on different test sets. Apart from these major findings, other minor but interesting insights are also discussed.

DiffuSeq Sequence to Sequence Text Generation with Diffusion Models ; Recently, diffusion models have emerged as a new paradigm for generative models. Despite the success in domains using continuous signals such as vision and audio, adapting diffusion models to natural language is underexplored due to the discrete nature of texts, especially for conditional generation. We tackle this challenge by proposing DiffuSeq a diffusion model designed for sequencetosequence Seq2Seq text generation tasks. Upon extensive evaluation over a wide range of Seq2Seq tasks, we find DiffuSeq achieving comparable or even better performance than six established baselines, including a stateoftheart model that is based on pretrained language models. Apart from quality, an intriguing property of DiffuSeq is its high diversity during generation, which is desired in many Seq2Seq tasks. We further include a theoretical analysis revealing the connection between DiffuSeq and autoregressivenonautoregressive models. Bringing together theoretical analysis and empirical evidence, we demonstrate the great potential of diffusion models in complex conditional language generation tasks. Code is available at urlhttpsgithub.comSharkNLPDiffuSeq

Automatic Generation of German Drama Texts Using Fine Tuned GPT2 Models ; This study is devoted to the automatic generation of German drama texts. We suggest an approach consisting of two key steps finetuning a GPT2 model the outline model to generate outlines of scenes based on keywords and finetuning a second model the generation model to generate scenes from the scene outline. The input for the neural model comprises two datasets the German Drama Corpus GerDraCor and German Text Archive Deutsches Textarchiv or DTA. In order to estimate the effectiveness of the proposed method, our models are compared with baseline GPT2 models. Our models perform well according to automatic quantitative evaluation, but, conversely, manual qualitative analysis reveals a poor quality of generated texts. This may be due to the quality of the dataset or training inputs.

Neural Artistic Style Transfer with Conditional Adversaria ; A neural artistic style transformation NST model can modify the appearance of a simple image by adding the style of a famous image. Even though the transformed images do not look precisely like artworks by the same artist of the respective style images, the generated images are appealing. Generally, a trained NST model specialises in a style, and a single image represents that style. However, generating an image under a new style is a tedious process, which includes full model training. In this paper, we present two methods that step toward the style image independent neural style transfer model. In other words, the trained model could generate semantically accurate generated image under any content, style image input pair. Our novel contribution is a unidirectionalGAN model that ensures the Cyclic consistency by the model architecture.Furthermore, this leads to much smaller model size and an efficient training and validation phase.

Understanding how Differentially Private Generative Models Spend their Privacy Budget ; Generative models trained with Differential Privacy DP are increasingly used to produce synthetic data while reducing privacy risks. Navigating their specific privacyutility tradeoffs makes it challenging to determine which models would work best for specific settingstasks. In this paper, we fill this gap in the context of tabular data by analyzing how DP generative models distribute privacy budgets across rows and columns, arguably the main source of utility degradation. We examine the main factors contributing to how privacy budgets are spent, including underlying modeling techniques, DP mechanisms, and data dimensionality. Our extensive evaluation of both graphical and deep generative models sheds light on the distinctive features that render them suitable for different settings and tasks. We show that graphical models distribute the privacy budget horizontally and thus cannot handle relatively wide datasets while the performance on the task they were optimized for monotonically increases with more data. Deep generative models spend their budget per iteration, so their behavior is less predictable with varying dataset dimensions but could perform better if trained on more features. Also, low levels of privacy epsilongeq100 could help some models generalize, achieving better results than without applying DP.

PaDGAN A Generative Adversarial Network for Performance Augmented Diverse Designs ; Deep generative models are proven to be a useful tool for automatic design synthesis and design space exploration. When applied in engineering design, existing generative models face three challenges 1 generated designs lack diversity and do not cover all areas of the design space, 2 it is difficult to explicitly improve the overall performance or quality of generated designs, and 3 existing models generally do not generate novel designs, outside the domain of the training data. In this paper, we simultaneously address these challenges by proposing a new Determinantal Point Processes based loss function for probabilistic modeling of diversity and quality. With this new loss function, we develop a variant of the Generative Adversarial Network, named Performance Augmented Diverse Generative Adversarial Network or PaDGAN, which can generate novel highquality designs with good coverage of the design space. Using three synthetic examples and one realworld airfoil design example, we demonstrate that PaDGAN can generate diverse and highquality designs. In comparison to a vanilla Generative Adversarial Network, on average, it generates samples with a 28 higher mean quality score with larger diversity and without the mode collapse issue. Unlike typical generative models that usually generate new designs by interpolating within the boundary of training data, we show that PaDGAN expands the design space boundary outside the training data towards highquality regions. The proposed method is broadly applicable to many tasks including design space exploration, design optimization, and creative solution recommendation.

Learning a powerful SVM using piecewise linear loss functions ; In this paper, we have considered general kpiecewise linear convex loss functions in SVM model for measuring the empirical risk. The resulting kPiecewise Linear loss Support Vector Machine kPLSVM model is an adaptive SVM model which can learn a suitable piecewise linear loss function according to nature of the given training set. The kPLSVM models are general SVM models and existing popular SVM models, like CSVM, LSSVM and PinSVM models, are their particular cases. We have performed the extensive numerical experiments with kPLSVM models for k 2 and 3 and shown that they are improvement over existing SVM models.

LSTM based Conversation Models ; In this paper, we present a conversational model that incorporates both context and participant role for twoparty conversations. Different architectures are explored for integrating participant role and context information into a Long Shortterm Memory LSTM language model. The conversational model can function as a language model or a language generation model. Experiments on the Ubuntu Dialog Corpus show that our model can capture multiple turn interaction between participants. The proposed method outperforms a traditional LSTM model as measured by language model perplexity and response ranking. Generated responses show characteristic differences between the two participant roles.

Ai4EComponentLib.jl A Componentbase Model Library in Julia ; Ai4EComponentLib.jlAi4EComponentLib is a componentbase model library based on Julia language, which relies on the differential equation solver DifferentialEquations.jl and the symbolic modeling tool Modelingtoolkit.jl. To handle problems in different physical domains, Ai4EComponentLib tries to build them with componentbase model. Supported by a new generation of symbolic modeling tools, models built with Ai4EComponentLib are more flexible and scalable than models built with traditional tools like Modelica. This paper will introduce the instance and general modeling methods of Ai4EComponentLib model library.

Generalization Metrics for Practical Quantum Advantage in Generative Models ; As the quantum computing community gravitates towards understanding the practical benefits of quantum computers, having a clear definition and evaluation scheme for assessing practical quantum advantage in the context of specific applications is paramount. Generative modeling, for example, is a widely accepted natural use case for quantum computers, and yet has lacked a concrete approach for quantifying success of quantum models over classical ones. In this work, we construct a simple and unambiguous approach to probe practical quantum advantage for generative modeling by measuring the algorithm's generalization performance. Using the samplebased approach proposed here, any generative model, from stateoftheart classical generative models such as GANs to quantum models such as Quantum Circuit Born Machines, can be evaluated on the same ground on a concrete welldefined framework. In contrast to other samplebased metrics for probing practical generalization, we leverage constrained optimization problems e.g., cardinalityconstrained problems and use these discrete datasets to define specific metrics capable of unambiguously measuring the quality of the samples and the model's generalization capabilities for generating data beyond the training set but still within the valid solution space. Additionally, our metrics can diagnose trainability issues such as mode collapse and overfitting, as we illustrate when comparing GANs to quantuminspired models built out of tensor networks. Our simulation results show that our quantuminspired models have up to a 68 times enhancement in generating unseen unique and valid samples compared to GANs, and a ratio of 612 for generating samples with better quality than those observed in the training set. We foresee these metrics as valuable tools for rigorously defining practical quantum advantage in the domain of generative modeling.

Generating Images Part by Part with Composite Generative Adversarial Networks ; Image generation remains a fundamental problem in artificial intelligence in general and deep learning in specific. The generative adversarial network GAN was successful in generating high quality samples of natural images. We propose a model called composite generative adversarial network, that reveals the complex structure of images with multiple generators in which each generator generates some part of the image. Those parts are combined by alpha blending process to create a new single image. It can generate, for example, background and face sequentially with two generators, after training on face dataset. Training was done in an unsupervised way without any labels about what each generator should generate. We found possibilities of learning the structure by using this generative model empirically.

Generalized HaldaneShastry Models as Supersymmetric Partners of the CalogeroSutherland Type Models ; We consider the supersymmetric CalogeroSutherland type Nparticle problems in one dimension and show that the corresponding fermionic part can be identified with the generalized XY models in the presence of an inhomogeneous magnetic field. In particular we show that the generalized HaldaneShastry models with magnetic field are themselves the fermionic partners of the CalogeroSutherland type models. Several such models are discussed and a recipe is given for constructing spin models and finding their ground state energy from the corresponding Nparticle problems.

On q Component Models on Cayley Tree The General Case ; In the paper we generalize results of paper 12 for a q component models on a Cayley tree of order kgeq 2. We generalize them in two directions 1 from k2 to any kgeq 2; 2 from concrete examples Potts and SOS models of q component models to any q component models with nearest neighbor interactions. We give a set of periodic ground states for the model. Using the contour argument which was developed in 12 we show existence of q different Gibbs measures for qcomponent models on Cayley tree of order kgeq 2.

Variable Selection and Model Averaging in Semiparametric Overdispersed Generalized Linear Models ; We express the mean and variance terms in a double exponential regression model as additive functions of the predictors and use Bayesian variable selection to determine which predictors enter the model, and whether they enter linearly or flexibly. When the variance term is null we obtain a generalized additive model, which becomes a generalized linear model if the predictors enter the mean linearly. The model is estimated using Markov chain Monte Carlo simulation and the methodology is illustrated using real and simulated data sets.

Modele FBSPPR des objets d'entreprise a la gestion dynamique des connaissances industrielles ; The phases of the life cycle of an industrial product can be described as a network of business processes. Products and informational materials are both raw materials and results of these processes. Modeling using generic model is a solution to integrate and value the enterprise and experts knowledge. Only a standardization approach involving several areas such as product modeling, process modeling, resource modeling and knowledge engineering can help build a retrieval system more efficient and profitable. The FunctionalBehaviorStructure approach is mix with the Product Process resources view in a global FBSPPRE generic model.

Zero Intelligence Models of the Continuous Double Auction Econometrics, Empirical Evidence and Generalization ; In the paper, a statistical procedure for estimating the parameters of zero intelligence models by means of tickbytick quote L1 data is proposed. A large class of existing zero intelligence models is reviewed. It is shown that all those models fail to describe the actual behavior of limit order books close to the ask price. A generalized model, accommodating the discrepancies found, is proposed and shown to give significant results for L1 data from three US electronic markets. It is also demonstrated that the generalized model preforms significantly better than the reviewed models.

Dynamic Entity Representations in Neural Language Models ; Understanding a long document requires tracking how entities are introduced and evolve over time. We present a new type of language model, EntityNLM, that can explicitly model entities, dynamically update their representations, and contextually generate their mentions. Our model is generative and flexible; it can model an arbitrary number of entities in context while generating each entity mention at an arbitrary length. In addition, it can be used for several different tasks such as language modeling, coreference resolution, and entity prediction. Experimental results with all these tasks demonstrate that our model consistently outperforms strong baselines and prior work.

A general class of mosaic random fields ; We present a model of a random field on a topological space M that unifies wellknown models such as the Poisson hyperplane tessellation model, the random token model, and the dead leaves model. In addition to generalizing these submodels from mathbbRd to other spaces such as the ddimensional unit sphere mathbbSd, our construction also extends the classical models themselves, e.g. by replacing the Poisson distribution by an arbitrary discrete distribution. Moreover, the method of construction directly produces an exact and fast simulation procedure. By investigating the covariance structure of the general model we recover various explicit correlation functions on mathbbRd and mathbbSd and obtain several new ones.

Generalizations of the Sommerfield and Schwinger models ; The Sommerfield model with a massive vector field coupled to a massless fermion in 11 dimensions is an exactly solvable analog of a BankZaks model. The physics of the model comprises a massive boson and an unparticle sector that survives at low energy as a conformal field theory Thirring model. We analyze generalizations of the Sommerfield model, and the corresponding generalizations of the Schwinger model, with more massless fermions and more vector fields.

On a generalized Kuramoto model with relativistic effects and emergent dynamics ; We propose a generalized Kuramoto model with relativistic effects and investigate emergent asymptotic behaviors. The proposed generalized Kuramoto model incorporates relativistic KuramotoRK type models which can be derived from the relativistic CuckerSmale RCS on the unit sphere under suitable approximations. We present several sufficient frameworks leading to complete synchronization in terms of initial data and system parameters. For the relativistic Kuramoto model, we show that it can be reduced to the Kuramoto model in any finite time interval in a nonrelativistic limit. We also provide several numerical examples for two approximations of the relativistic Kuramoto model, and compare them with analytical results.

Masked Adversarial Generation for Neural Machine Translation ; Attacking Neural Machine Translation models is an inherently combinatorial task on discrete sequences, solved with approximate heuristics. Most methods use the gradient to attack the model on each sample independently. Instead of mechanically applying the gradient, could we learn to produce meaningful adversarial attacks In contrast to existing approaches, we learn to attack a model by training an adversarial generator based on a language model. We propose the Masked Adversarial Generation MAG model, that learns to perturb the translation model throughout the training process. The experiments show that it improves the robustness of machine translation models, while being faster than competing methods.

BERTopic Neural topic modeling with a classbased TFIDF procedure ; Topic models can be useful tools to discover latent topics in collections of documents. Recent studies have shown the feasibility of approach topic modeling as a clustering task. We present BERTopic, a topic model that extends this process by extracting coherent topic representation through the development of a classbased variation of TFIDF. More specifically, BERTopic generates document embedding with pretrained transformerbased language models, clusters these embeddings, and finally, generates topic representations with the classbased TFIDF procedure. BERTopic generates coherent topics and remains competitive across a variety of benchmarks involving classical models and those that follow the more recent clustering approach of topic modeling.

LANCE Stresstesting Visual Models by Generating Languageguided Counterfactual Images ; We propose an automated algorithm to stresstest a trained visual model by generating languageguided counterfactual test images LANCE. Our method leverages recent progress in large language modeling and textbased image editing to augment an IID test set with a suite of diverse, realistic, and challenging test images without altering model weights. We benchmark the performance of a diverse set of pretrained models on our generated data and observe significant and consistent performance drops. We further analyze model sensitivity across different types of edits, and demonstrate its applicability at surfacing previously unknown classlevel model biases in ImageNet.

Diluted Generalized Random Energy Model ; We introduce a layered random spin model, equivalent to the Generalized Random Energy Model GREM. In analogy with diluted spin systems, a diluted GREM DGREM is introduced.It can be applied to calculate approximately thermodynamic properties of spin glass models in low dimensions. For Edwards Anderson model it gives correct critical dimension and 5 accuracy for ground state energy in 2d.

Exact ground state of the generalized threedimensional ShastrySutherland model ; We generalize the ShastrySutherland model to three dimensions. By representing the model as a sum of the semidefinite positive projection operators, we exactly prove that the model has exact dimer ground state. Several schemes for constructing the threedimensional ShastrySutherland model are proposed.

Generalized Cubic Model for BaTiO3like Ferroelectric Substance ; We propose an orderdisorder type microscopic model for BaTiO3like Ferroelectric Substance. Our model has three phase transitions and four phases. The symmetry and directions of the polarizations of the ordered phases agree with the experimental results of BaTiO3. The intermediate phases in our model are known as an incompletely ordered phase, which appears in a generalized clock model.

Duality of a Generalized Gauge Invariant Ising Model on Random Surfaces ; A generalized gauge invariant Ising model on random surfaces with nontrivial topology is proposed and investigated with the dual transformation. It is proved that the model is selfdual in case of a selfdual lattice. In special cases the model reduces to the known solvable Isingtype models.

Matrix Models for Beta Ensembles ; This paper constructs tridiagonal random matrix models for general beta0 betaHermite Gaussian and betaLaguerre Wishart ensembles. These generalize the wellknown Gaussian and Wishart models for beta 1,2,4. Furthermore, in the cases of the betaLaguerre ensembles, we eliminate the exponent quantization present in the previously known models. We further discuss applications for the new matrix models, and present some open problems.

Discrete mechanics a kinematics for a particular case of causal sets ; The model is a particular case of causal set. This is a discrete model of spacetime in a microscopic level. In paper the most general properties of the model are investigated without any reference to a dynamics. The dynamics of the model is introduced in arXiv 1004.5077. These two papers introduce a consistent description of the model.

Nonrelativistic matter and Dark energy in a quantum conformal model ; We consider a generalization of the standard model which respects quantum conformal invariance. This model leads to identically zero vacuum energy. We show how nonrelativistic matter and dark energy arises in this model. Hence the model is shown to be consistent with observations.

Computational Models for Multiview Dense Depth Maps of Dynamic Scene ; This paper reviews the recent progresses of the depth map generation for dynamic scene and its corresponding computational models. This paper mainly covers the homogeneous ambiguity models in depth sensing, resolution models in depth processing, and consistency models in depth optimization. We also summarize the future work in the depth map generation.

Model Averaging for Generalized Linear Model with Covariates that are Missing completely at Random ; In this paper, we consider the estimation of generalized linear models with covariates that are missing completely at random. We propose a model averaging estimation method and prove that the corresponding model averaging estimator is asymptotically optimal under certain assumptions. Simulaiton results illustrate that this method has better performance than other alternatives under most situations.

The multitrace matrix model An alternative to Connes NCG and IKKT model ; We present a new multitrace matrix model, which is a generalization of the real quartic one matrix model, exhibiting dynamical emergence of a fuzzy twosphere and its noncommutative gauge theory. This provides a novel and a much simpler alternative to Connes noncommutative geometry and to the IKKT matrix model for emergent geometry in two dimensions.

Supplemental Material Lifelong Generative Modelling Using Dynamic Expansion Graph Model ; In this article, we provide the appendix for Lifelong Generative Modelling Using Dynamic Expansion Graph Model. This appendix includes additional visual results as well as the numerical results on the challenging datasets. In addition, we also provide detailed proofs for the proposed theoretical analysis framework. The source code can be found in httpsgithub.comdtuzi123ExpansionGraphModel.

Investigating Memorization of Conspiracy Theories in Text Generation ; The adoption of natural language generation NLG models can leave individuals vulnerable to the generation of harmful information memorized by the models, such as conspiracy theories. While previous studies examine conspiracy theories in the context of social media, they have not evaluated their presence in the new space of generative language models. In this work, we investigate the capability of language models to generate conspiracy theory text. Specifically, we aim to answer can we test pretrained generative language models for the memorization and elicitation of conspiracy theories without access to the model's training data We highlight the difficulties of this task and discuss it in the context of memorization, generalization, and hallucination. Utilizing a new dataset consisting of conspiracy theory topics and machinegenerated conspiracy theories helps us discover that many conspiracy theories are deeply rooted in the pretrained language models. Our experiments demonstrate a relationship between model parameters such as size and temperature and their propensity to generate conspiracy theory text. These results indicate the need for a more thorough review of NLG applications before release and an indepth discussion of the drawbacks of memorization in generative language models.

Crystal Transformer Selflearning neural language model for Generative and Tinkering Design of Materials ; Selfsupervised neural language models have recently achieved unprecedented success, from natural language processing to learning the languages of biological sequences and organic molecules. These models have demonstrated superior performance in the generation, structure classification, and functional predictions for proteins and molecules with learned representations. However, most of the maskingbased pretrained language models are not designed for generative design, and their blackbox nature makes it difficult to interpret their design logic. Here we propose BLMM Crystal Transformer, a neural network based probabilistic generative model for generative and tinkering design of inorganic materials. Our model is built on the blank filling language model for text generation and has demonstrated unique advantages in learning the materials grammars together with highquality generation, interpretability, and data efficiency. It can generate chemically valid materials compositions with as high as 89.7 charge neutrality and 84.8 balanced electronegativity, which are more than 4 and 8 times higher compared to a pseudo random sampling baseline. The probabilistic generation process of BLMM allows it to recommend tinkering operations based on learned materials chemistry and makes it useful for materials doping. Combined with the TCSP crysal structure prediction algorithm, We have applied our model to discover a set of new materials as validated using DFT calculations. Our work thus brings the unsupervised transformer language models based generative artificial intelligence to inorganic materials. A userfriendly web app has been developed for computational materials doping and can be accessed freely at urlwww.materialsatlas.orgblmtinker.

ToolCoder Teach Code Generation Models to use API search tools ; Automatically generating source code from natural language descriptions has been a growing field of research in recent years. However, current largescale code generation models often encounter difficulties when selecting appropriate APIs for specific contexts. These models may generate APIs that do not meet requirements or refer to nonexistent APIs in thirdparty libraries, especially for lesserknown or private libraries. Inspired by the process of human developers using tools to search APIs, we propose ToolCoder, a novel approach that integrates API search tools with existing models to assist in code generation and API selection. To teach our model to use tools, we introduce an automated data annotation method using ChatGPT to add tool usage information into the source code data and finetune code generation models. During inference, we integrate API search tools into the generation process so that our model can automatically use the search tool to get suggestions when selecting an API. Our experimental results demonstrate that ToolCoder exhibits excellent performance and generalization across five public and private library code generation benchmarks, with at least 6.21 improvement on average pass1 metrics and 9.64 improvement on average pass10 metrics compared to stateoftheart methods. Furthermore, we show that our relatively small ToolCoder model is comparable to one of the current best models, GPT3.5, highlighting the potential of incorporating programming tools into the code generation process.

Interactive Fashion Content Generation Using LLMs and Latent Diffusion Models ; Fashionable image generation aims to synthesize images of diverse fashion prevalent around the globe, helping fashion designers in realtime visualization by giving them a basic customized structure of how a specific design preference would look in real life and what further improvements can be made for enhanced customer satisfaction. Moreover, users can alone interact and generate fashionable images by just giving a few simple prompts. Recently, diffusion models have gained popularity as generative models owing to their flexibility and generation of realistic images from Gaussian noise. Latent diffusion models are a type of generative model that use diffusion processes to model the generation of complex data, such as images, audio, or text. They are called latent because they learn a hidden representation, or latent variable, of the data that captures its underlying structure. We propose a method exploiting the equivalence between diffusion models and energybased models EBMs and suggesting ways to compose multiple probability distributions. We describe a pipeline on how our method can be used specifically for new fashionable outfit generation and virtual tryon using LLMguided texttoimage generation. Our results indicate that using an LLM to refine the prompts to the latent diffusion model assists in generating globally creative and culturally diversified fashion styles and reducing bias.

VideoControlNet A MotionGuided VideotoVideo Translation Framework by Using Diffusion Model with ControlNet ; Recently, diffusion models like StableDiffusion have achieved impressive image generation results. However, the generation process of such diffusion models is uncontrollable, which makes it hard to generate videos with continuous and consistent content. In this work, by using the diffusion model with ControlNet, we proposed a new motionguided videotovideo translation framework called VideoControlNet to generate various videos based on the given prompts and the condition from the input video. Inspired by the video codecs that use motion information for reducing temporal redundancy, our framework uses motion information to prevent the regeneration of the redundant areas for content consistency. Specifically, we generate the first frame i.e., the Iframe by using the diffusion model with ControlNet. Then we generate other key frames i.e., the Pframe based on the previous IPframe by using our newly proposed motionguided Pframe generation MgPG method, in which the Pframes are generated based on the motion information and the occlusion areas are inpainted by using the diffusion model. Finally, the rest frames i.e., the Bframe are generated by using our motionguided Bframe interpolation MgBI module. Our experiments demonstrate that our proposed VideoControlNet inherits the generation capability of the pretrained large diffusion model and extends the image diffusion model to the video diffusion model by using motion information. More results are provided at our project page.

PluGeN MultiLabel Conditional Generation From PreTrained Models ; Modern generative models achieve excellent quality in a variety of tasks including image or text generation and chemical molecule modeling. However, existing methods often lack the essential ability to generate examples with requested properties, such as the age of the person in the photo or the weight of the generated molecule. Incorporating such additional conditioning factors would require rebuilding the entire architecture and optimizing the parameters from scratch. Moreover, it is difficult to disentangle selected attributes so that to perform edits of only one attribute while leaving the others unchanged. To overcome these limitations we propose PluGeN Plugin Generative Network, a simple yet effective generative technique that can be used as a plugin to pretrained generative models. The idea behind our approach is to transform the entangled latent representation using a flowbased module into a multidimensional space where the values of each attribute are modeled as an independent onedimensional distribution. In consequence, PluGeN can generate new samples with desired attributes as well as manipulate labeled attributes of existing examples. Due to the disentangling of the latent representation, we are even able to generate samples with rare or unseen combinations of attributes in the dataset, such as a young person with gray hair, men with makeup, or women with beards. We combined PluGeN with GAN and VAE models and applied it to conditional generation and manipulation of images and chemical molecule modeling. Experiments demonstrate that PluGeN preserves the quality of backbone models while adding the ability to control the values of labeled attributes.

MolHF A Hierarchical Normalizing Flow for Molecular Graph Generation ; Molecular de novo design is a critical yet challenging task in scientific fields, aiming to design novel molecular structures with desired property profiles. Significant progress has been made by resorting to generative models for graphs. However, limited attention is paid to hierarchical generative models, which can exploit the inherent hierarchical structure with rich semantic information of the molecular graphs and generate complex molecules of larger size that we shall demonstrate to be difficult for most existing models. The primary challenge to hierarchical generation is the nondifferentiable issue caused by the generation of intermediate discrete coarsened graph structures. To sidestep this issue, we cast the tricky hierarchical generation problem over discrete spaces as the reverse process of hierarchical representation learning and propose MolHF, a new hierarchical flowbased model that generates molecular graphs in a coarsetofine manner. Specifically, MolHF first generates bonds through a multiscale architecture, then generates atoms based on the coarsened graph structure at each scale. We demonstrate that MolHF achieves stateoftheart performance in random generation and property optimization, implying its high capacity to model data distribution. Furthermore, MolHF is the first flowbased model that can be applied to model larger molecules polymer with more than 100 heavy atoms. The code and models are available at httpsgithub.comvioletstoMolHF.

PatternGPT A PatternDriven Framework for Large Language Model Text Generation ; Large language modelsLLMShave shown excellent text generation capabilities, capable of generating fluent humanlike responses for many downstream tasks. However, applying large language models to realworld critical tasks remains challenging due to their susceptibility to hallucinations and inability to directly use external knowledge. To cope with the above challenges, this paper proposes PatternGPT, a patterndriven text generation framework for Large Language Models. Firstly, the framework utilizes the extraction capability of Large Language Models to generate rich and diversified structured and formalized patterns, which facilitates the introduction of external knowledge to do the computation, and then draws on the idea of federated learning to use multiple agents to achieve the sharing in order to obtain more diversified patterns, and finally uses judgment criteria and optimization algorithm to search for highquality patterns to guide the generation of models. Finally, external knowledge such as judgment criteria and optimization algorithms are used to search for highquality patterns, and the searched patterns are used to guide model generation. This framework has the advantages of generating diversified patterns, protecting data privacy, combining external knowledge, and improving the quality of generation, which provides an effective method to optimize the text generation capability of large language models, and make it better applied to the field of intelligent dialogue and content generation.

Steered Diffusion A Generalized Framework for PlugandPlay Conditional Image Synthesis ; Conditional generative models typically demand large annotated training sets to achieve highquality synthesis. As a result, there has been significant interest in designing models that perform plugandplay generation, i.e., to use a predefined or pretrained model, which is not explicitly trained on the generative task, to guide the generative process e.g., using language. However, such guidance is typically useful only towards synthesizing highlevel semantics rather than editing finegrained details as in imagetoimage translation tasks. To this end, and capitalizing on the powerful finegrained generative control offered by the recent diffusionbased generative models, we introduce Steered Diffusion, a generalized framework for photorealistic zeroshot conditional image generation using a diffusion model trained for unconditional generation. The key idea is to steer the image generation of the diffusion model at inference time via designing a loss using a pretrained inverse model that characterizes the conditional task. This loss modulates the sampling trajectory of the diffusion process. Our framework allows for easy incorporation of multiple conditions during inference. We present experiments using steered diffusion on several tasks including inpainting, colorization, textguided semantic editing, and image superresolution. Our results demonstrate clear qualitative and quantitative improvements over stateoftheart diffusionbased plugandplay models while adding negligible additional computational cost.

On Hierarchical MultiResolution Graph Generative Models ; In real world domains, most graphs naturally exhibit a hierarchical structure. However, datadriven graph generation is yet to effectively capture such structures. To address this, we propose a novel approach that recursively generates community structures at multiple resolutions, with the generated structures conforming to training data distribution at each level of the hierarchy. The graphs generation is designed as a sequence of coarsetofine generative models allowing for parallel generation of all substructures, resulting in a high degree of scalability. Our method demonstrates generative performance improvement on multiple graph datasets.

FFPDG Fast, Fair and Private Data Generation ; Generative modeling has been used frequently in synthetic data generation. Fairness and privacy are two big concerns for synthetic data. Although Recent GAN citegoodfellow2014generative based methods show good results in preserving privacy, the generated data may be more biased. At the same time, these methods require high computation resources. In this work, we design a fast, fair, flexible and private data generation method. We show the effectiveness of our method theoretically and empirically. We show that models trained on data generated by the proposed method can perform well in inference stage on real application scenarios.

Classical and Quantum Intertwine ; Model interactions between classical and quantum systems are briefly discussed. These include general measurementlike couplings, SternGerlach experiment, model of a counter, quantum Zeno effect, SQUIDtank model.

Mathematical Model of Shock Waves ; Presented here is the mathematical model describing the phenomenon of shock waves. The underlying concept is based on the timespace model of wave propagation.

Using BuiltIn DomainSpecific Modeling Support to Guide ModelBased Test Generation ; We present a modelbased testing approach to support automated test generation with domainspecific concepts. This includes a language expert who is an expert at building test models and domain experts who are experts in the domain of the system under test. First, we provide a framework to support the language expert in building test models using a full Java programming language with the help of simple but powerful modeling elements of the framework. Second, based on the model built with this framework, the toolset automatically forms a domainspecific modeling language that can be used to further constrain and guide test generation from these models by a domain expert. This makes it possible to generate a large set of test cases covering the full model, chosen constrained parts of the model, or manually define specific test cases on top of the model while using concepts familiar to the domain experts.

Smoothing parameter and model selection for general smooth models ; This paper discusses a general framework for smoothing parameter estimation for models with regular likelihoods constructed in terms of unknown smooth functions of covariates. Gaussian random effects and parametric terms may also be present. By construction the method is numerically stable and convergent, and enables smoothing parameter uncertainty to be quantified. The latter enables us to fix a well known problem with AIC for such models. The smooth functions are represented by reduced rank spline like smoothers, with associated quadratic penalties measuring function smoothness. Model estimation is by penalized likelihood maximization, where the smoothing parameters controlling the extent of penalization are estimated by Laplace approximate marginal likelihood. The methods cover, for example, generalized additive models for nonexponential family responses for example beta, ordered categorical, scaled t distribution, negative binomial and Tweedie distributions, generalized additive models for location scale and shape for example two stage zero inflation models, and Gaussian locationscale models, Cox proportional hazards models and multivariate additive models. The framework reduces the implementation of new model classes to the coding of some standard derivatives of the log likelihood.

Modular and Incremental Global Model Management with Extended Generalized Discrimination Networks ; Complex projects developed under the paradigm of modeldriven engineering nowadays often involve several interrelated models, which are automatically processed via a multitude of model operations. Modular and incremental construction and execution of such networks of models and model operations are required to accommodate efficient development with potentially largescale models. The underlying problem is also called Global Model Management. In this report, we propose an approach to modular and incremental Global Model Management via an extension to the existing technique of Generalized Discrimination Networks GDNs. In addition to further generalizing the notion of query operations employed in GDNs, we adapt the previously queryonly mechanism to operations with side effects to integrate model transformation and model synchronization. We provide incremental algorithms for the execution of the resulting extended Generalized Discrimination Networks eGDNs, as well as a prototypical implementation for a number of example eGDN operations. Based on this prototypical implementation, we experiment with an application scenario from the software development domain to empirically evaluate our approach with respect to scalability and conceptually demonstrate its applicability in a typical scenario. Initial results confirm that the presented approach can indeed be employed to realize efficient Global Model Management in the considered scenario.

Generative AgentBased Modeling Unveiling Social System Dynamics through Coupling Mechanistic Models with Generative Artificial Intelligence ; We discuss the emerging new opportunity for building feedbackrich computational models of social systems using generative artificial intelligence. Referred to as Generative AgentBased Models GABMs, such individuallevel models utilize large language models such as ChatGPT to represent human decisionmaking in social settings. We provide a GABM case in which human behavior can be incorporated in simulation models by coupling a mechanistic model of human interactions with a pretrained large language model. This is achieved by introducing a simple GABM of social norm diffusion in an organization. For educational purposes, the model is intentionally kept simple. We examine a wide range of scenarios and the sensitivity of the results to several changes in the prompt. We hope the article and the model serve as a guide for building useful diffusion models that include realistic human reasoning and decisionmaking.

Tensor models with generalized melonic interactions ; Tensor models are natural generalizations of matrix models. The interactions and observables in the case of unitary invariant models are generalizations of matrix traces. Some notable interactions in the literature include the melonic ones, the tetrahedral one as well as the planar ones in rank three, or necklaces in even ranks. Here we introduce generalized melonic interactions which generalize the melonic and necklace interactions. We characterize them as treelike gluings of quartic interactions. We also completely characterize the Feynman graphs which contribute to the large N limit. For a subclass of generalized melonic interactions called totally unbalanced interactions, we prove that the large N limit is Gaussian and therefore the Feynman graphs are in bijection with trees. This result further extends the class of tensor models which fall into the Gaussian universality class. Another key aspect of tensor models with generalized melonic interactions is that they can be written as matrix models without increasing the number of degrees of freedom of the original tensor models. In the case of totally unbalanced interactions, this new matrix model formulation in fact decreases the number of degrees of freedom, meaning that some of the original degrees of freedom are effectively integrated. We then show how the large N Gaussian behavior can be reproduced using a saddle point analysis on those matrix models.

A Generative Approach for Mitigating Structural Biases in Natural Language Inference ; Many natural language inference NLI datasets contain biases that allow models to perform well by only using a biased subset of the input, without considering the remainder features. For instance, models are able to make a classification decision by only using the hypothesis, without learning the true relationship between it and the premise. These structural biases lead discriminative models to learn unintended superficial features and to generalize poorly out of the training distribution. In this work, we reformulate the NLI task as a generative task, where a model is conditioned on the biased subset of the input and the label and generates the remaining subset of the input. We show that by imposing a uniform prior, we obtain a provably unbiased model. Through synthetic experiments, we find that this approach is highly robust to large amounts of bias. We then demonstrate empirically on two types of natural bias that this approach leads to fully unbiased models in practice. However, we find that generative models are difficult to train and they generally perform worse than discriminative baselines. We highlight the difficulty of the generative modeling task in the context of NLI as a cause for this worse performance. Finally, by finetuning the generative model with a discriminative objective, we reduce the performance gap between the generative model and the discriminative baseline, while allowing for a small amount of bias.

Learning BodyAware 3D Shape Generative Models ; The shape of many objects in the built environment is dictated by their relationships to the human body how will a person interact with this object Existing datadriven generative models of 3D shapes produce plausible objects but do not reason about the relationship of those objects to the human body. In this paper, we learn bodyaware generative models of 3D shapes. Specifically, we train generative models of chairs, an ubiquitous shape category, which can be conditioned on a given body shape or sitting pose. The bodyshapeconditioned models produce chairs which will be comfortable for a person with the given body shape; the poseconditioned models produce chairs which accommodate the given sitting pose. To train these models, we define a sitting pose matching metric and a novel sitting comfort metric. Calculating these metrics requires an expensive optimization to sit the body into the chair, which is too slow to be used as a loss function for training a generative model. Thus, we train neural networks to efficiently approximate these metrics. We use our approach to train three bodyaware generative shape models a structured partbased generator, a point cloud generator, and an implicit surface generator. In all cases, our approach produces models which adapt their output chair shapes to input human body specifications.

PreTrained Neural Language Models for Automatic Mobile App User Feedback Answer Generation ; Studies show that developers' answers to the mobile app users' feedbacks on app stores can increase the apps' star rating. To help app developers generate answers that are related to the users' issues, recent studies develop models to generate the answers automatically. Aims The app response generation models use deep neural networks and require training data. PreTrained neural language Models PTM used in Natural Language Processing NLP take advantage of the information they learned from a large corpora in an unsupervised manner, and can reduce the amount of required training data. In this paper, we evaluate PTMs to generate replies to the mobile app user feedbacks. Method We train a Transformer model from scratch and finetune two PTMs to evaluate the generated responses, which are compared to RRGEN, a current app response model. We also evaluate the models with different portions of the training data. Results The results on a large dataset evaluated by automatic metrics show that PTMs obtain lower scores than the baselines. However, our human evaluation confirms that PTMs can generate more relevant and meaningful responses to the posted feedbacks. Moreover, the performance of PTMs has less drop compared to other models when the amount of training data is reduced to 13. Conclusion PTMs are useful in generating responses to app reviews and are more robust models to the amount of training data provided. However, the prediction time is 19X than RRGEN. This study can provide new avenues for research in adapting the PTMs for analyzing mobile app user feedbacks. Index Termsmobile app user feedback analysis, neural pretrained language models, automatic answer generation

Evaluation of Categorical Generative Models Bridging the Gap Between Real and Synthetic Data ; The machine learning community has mainly relied on real data to benchmark algorithms as it provides compelling evidence of model applicability. Evaluation on synthetic datasets can be a powerful tool to provide a better understanding of a model's strengths, weaknesses, and overall capabilities. Gaining these insights can be particularly important for generative modeling as the target quantity is completely unknown. Multiple issues related to the evaluation of generative models have been reported in the literature. We argue those problems can be avoided by an evaluation based on ground truth. General criticisms of synthetic experiments are that they are too simplified and not representative of practical scenarios. As such, our experimental setting is tailored to a realistic generative task. We focus on categorical data and introduce an appropriately scalable evaluation method. Our method involves tasking a generative model to learn a distribution in a highdimensional setting. We then successively bin the large space to obtain smaller probability spaces where meaningful statistical tests can be applied. We consider increasingly large probability spaces, which correspond to increasingly difficult modeling tasks and compare the generative models based on the highest task difficulty they can reach before being detected as being too far from the ground truth. We validate our evaluation procedure with synthetic experiments on both synthetic generative models and current stateoftheart categorical generative models.

Spot the fake lungs Generating Synthetic Medical Images using Neural Diffusion Models ; Generative models are becoming popular for the synthesis of medical images. Recently, neural diffusion models have demonstrated the potential to generate photorealistic images of objects. However, their potential to generate medical images is not explored yet. In this work, we explore the possibilities of synthesis of medical images using neural diffusion models. First, we use a pretrained DALLE2 model to generate lungs XRay and CT images from an input text prompt. Second, we train a stable diffusion model with 3165 XRay images and generate synthetic images. We evaluate the synthetic image data through a qualitative analysis where two independent radiologists label randomly chosen samples from the generated data as real, fake, or unsure. Results demonstrate that images generated with the diffusion model can translate characteristics that are otherwise very specific to certain medical conditions in chest XRay or CT images. Careful tuning of the model can be very promising. To the best of our knowledge, this is the first attempt to generate lungs XRay and CT images using neural diffusion models. This work aims to introduce a new dimension in artificial intelligence for medical imaging. Given that this is a new topic, the paper will serve as an introduction and motivation for the research community to explore the potential of diffusion models for medical image synthesis. We have released the synthetic images on httpswww.kaggle.comdatasetshazratawesomelungs.

Textile Pattern Generation Using Diffusion Models ; The problem of textguided image generation is a complex task in Computer Vision, with various applications, including creating visually appealing artwork and realistic product images. One popular solution widely used for this task is the diffusion model, a generative model that generates images through an iterative process. Although diffusion models have demonstrated promising results for various image generation tasks, they may only sometimes produce satisfactory results when applied to more specific domains, such as the generation of textile patterns based on text guidance. This study presents a finetuned diffusion model specifically trained for textile pattern generation by text guidance to address this issue. The study involves the collection of various textile pattern images and their captioning with the help of another AI model. The finetuned diffusion model is trained with this newly created dataset, and its results are compared with the baseline models visually and numerically. The results demonstrate that the proposed finetuned diffusion model outperforms the baseline models in terms of pattern quality and efficiency in textile pattern generation by text guidance. This study presents a promising solution to the problem of textguided textile pattern generation and has the potential to simplify the design process within the textile industry.

Comparative Assessment of Markov Models and Recurrent Neural Networks for Jazz Music Generation ; As generative models have risen in popularity, a domain that has risen alongside is generative models for music. Our study aims to compare the performance of a simple Markov chain model and a recurrent neural network RNN model, two popular models for sequence generating tasks, in jazz music improvisation. While music, especially jazz, remains subjective in telling whether a composition is good or bad, we aim to quantify our results using metrics of groove pattern similarity and pitch class histogram entropy. We trained both models using transcriptions of jazz blues choruses from professional jazz players, and also fed musical jazz seeds to help give our model some context in beginning the generation. Our results show that the RNN outperforms the Markov model on both of our metrics, indicating better rhythmic consistency and tonal stability in the generated music. Through the use of music21 library, we tokenized our jazz dataset into pitches and durations that our model could interpret and train on. Our findings contribute to the growing field of AIgenerated music, highlighting the important use of metrics to assess generation quality. Future work includes expanding the dataset of MIDI files to a larger scale, conducting human surveys for subjective evaluations, and incorporating additional metrics to address the challenge of subjectivity in music evaluation. Our study provides valuable insight into the use of recurrent neural networks for sequential based tasks like generating music.

ObjectiveReinforced Generative Adversarial Networks ORGAN for Sequence Generation Models ; In unsupervised data generation tasks, besides the generation of a sample based on previous observations, one would often like to give hints to the model in order to bias the generation towards desirable metrics. We propose a method that combines Generative Adversarial Networks GANs and reinforcement learning RL in order to accomplish exactly that. While RL biases the data generation process towards arbitrary metrics, the GAN component of the reward function ensures that the model still remembers information learned from data. We build upon previous results that incorporated GANs and RL in order to generate sequence data and test this model in several settings for the generation of molecules encoded as text sequences SMILES and in the context of music generation, showing for each case that we can effectively bias the generation process towards desired metrics.

A Deep Generative Framework for Paraphrase Generation ; Paraphrase generation is an important problem in NLP, especially in question answering, information retrieval, information extraction, conversation systems, to name a few. In this paper, we address the problem of generating paraphrases automatically. Our proposed method is based on a combination of deep generative models VAE with sequencetosequence models LSTM to generate paraphrases, given an input sentence. Traditional VAEs when combined with recurrent neural networks can generate free text but they are not suitable for paraphrase generation for a given sentence. We address this problem by conditioning the both, encoder and decoder sides of VAE, on the original sentence, so that it can generate the given sentence's paraphrases. Unlike most existing models, our model is simple, modular and can generate multiple paraphrases, for a given sentence. Quantitative evaluation of the proposed method on a benchmark paraphrase dataset demonstrates its efficacy, and its performance improvement over the stateoftheart methods by a significant margin, whereas qualitative human evaluation indicate that the generated paraphrases are wellformed, grammatically correct, and are relevant to the input sentence. Furthermore, we evaluate our method on a newly released question paraphrase dataset, and establish a new baseline for future research.

Manifoldvalued Image Generation with Wasserstein Generative Adversarial Nets ; Generative modeling over natural images is one of the most fundamental machine learning problems. However, few modern generative models, including Wasserstein Generative Adversarial Nets WGANs, are studied on manifoldvalued images that are frequently encountered in realworld applications. To fill the gap, this paper first formulates the problem of generating manifoldvalued images and exploits three typical instances huesaturationvalue HSV color image generation, chromaticitybrightness CB color image generation, and diffusiontensor DT image generation. For the proposed generative modeling problem, we then introduce a theorem of optimal transport to derive a new Wasserstein distance of data distributions on complete manifolds, enabling us to achieve a tractable objective under the WGAN framework. In addition, we recommend three benchmark datasets that are CIFAR10 HSVCB color images, ImageNet HSVCB color images, UCL DT image datasets. On the three datasets, we experimentally demonstrate the proposed manifoldaware WGAN model can generate more plausible manifoldvalued images than its competitors.

Generative chemistry drug discovery with deep learning generative models ; The de novo design of molecular structures using deep learning generative models introduces an encouraging solution to drug discovery in the face of the continuously increased cost of new drug development. From the generation of original texts, images, and videos, to the scratching of novel molecular structures, the incredible creativity of deep learning generative models surprised us about the height machine intelligence can achieve. The purpose of this paper is to review the latest advances in generative chemistry which relies on generative modeling to expedite the drug discovery process. This review starts with a brief history of artificial intelligence in drug discovery to outline this emerging paradigm. Commonly used chemical databases, molecular representations, and tools in cheminformatics and machine learning are covered as the infrastructure for the generative chemistry. The detailed discussions on utilizing cuttingedge generative architectures, including recurrent neural network, variational autoencoder, adversarial autoencoder, and generative adversarial network for compound generation are focused. Challenges and future perspectives follow.

Deformable Generator Network Unsupervised Disentanglement of Appearance and Geometry ; We present a deformable generator model to disentangle the appearance and geometric information for both image and video data in a purely unsupervised manner. The appearance generator network models the information related to appearance, including color, illumination, identity or category, while the geometric generator performs geometric warping, such as rotation and stretching, through generating deformation field which is used to warp the generated appearance to obtain the final image or video sequences. Two generators take independent latent vectors as input to disentangle the appearance and geometric information from image or video sequences. For video data, a nonlinear transition model is introduced to both the appearance and geometric generators to capture the dynamics over time. The proposed scheme is general and can be easily integrated into different generative models. An extensive set of qualitative and quantitative experiments shows that the appearance and geometric information can be well disentangled, and the learned geometric generator can be conveniently transferred to other image datasets to facilitate knowledge transfer tasks.

EnsembleGAN Adversarial Learning for RetrievalGeneration Ensemble Model on ShortText Conversation ; Generating qualitative responses has always been a challenge for humancomputer dialogue systems. Existing dialogue systems generally derive from either retrievalbased or generativebased approaches, both of which have their own pros and cons. Despite the natural idea of an ensemble model of the two, existing ensemble methods only focused on leveraging one approach to enhance another, we argue however that they can be further mutually enhanced with a proper training strategy. In this paper, we propose ensembleGAN, an adversarial learning framework for enhancing a retrievalgeneration ensemble model in opendomain conversation scenario. It consists of a languagemodellike generator, a ranker generator, and one ranker discriminator. Aiming at generating responses that approximate the groundtruth and receive high ranking scores from the discriminator, the two generators learn to generate improved highly relevant responses and competitive unobserved candidates respectively, while the discriminative ranker is trained to identify true responses from adversarial ones, thus featuring the merits of both generator counterparts. The experimental results on a large shorttext conversation data demonstrate the effectiveness of the ensembleGAN by the amelioration on both human and automatic evaluation metrics.

Interpreting Spatially Infinite Generative Models ; Traditional deep generative models of images and other spatial modalities can only generate fixed sized outputs. The generated images have exactly the same resolution as the training images, which is dictated by the number of layers in the underlying neural network. Recent work has shown, however, that feeding spatial noise vectors into a fully convolutional neural network enables both generation of arbitrary resolution output images as well as training on arbitrary resolution training images. While this work has provided impressive empirical results, little theoretical interpretation was provided to explain the underlying generative process. In this paper we provide a firm theoretical interpretation for infinite spatial generation, by drawing connections to spatial stochastic processes. We use the resulting intuition to improve upon existing spatially infinite generative models to enable more efficient training through a model that we call an infinite generative adversarial network, or inftyGAN. Experiments on world map generation, panoramic images and texture synthesis verify the ability of inftyGAN to efficiently generate images of arbitrary size.

Generation of nonstationary stochastic fields using Generative Adversarial Networks ; In the context of generating geological facies conditioned on observed data, samples corresponding to all possible conditions are not generally available in the training set and hence the generation of these realizations depends primary on the generalization capability of the trained generative model. The problem becomes more complex when applied on nonstationary fields. In this work, we investigate the problem of using Generative Adversarial Networks GANs models to generate nonstationary geological channelized patterns and examine the models generalization capability at new spatial modes that were never seen in the given training set. The developed training method based on spatialconditioning allowed for effective learning of the correlation between the spatial conditions i.e. nonstationary maps and the realizations implicitly without using additional loss terms or solving optimization problems for every new given data after training. In addition, our models can be trained on 2D and 3D samples. The results on real and artificial datasets show that we were able to generate geologicallyplausible realizations beyond the training samples and with a strong correlation with the target maps.

Factual and Informative Review Generation for Explainable Recommendation ; Recent models can generate fluent and grammatical synthetic reviews while accurately predicting user ratings. The generated reviews, expressing users' estimated opinions towards related products, are often viewed as natural language 'rationales' for the jointly predicted rating. However, previous studies found that existing models often generate repetitive, universally applicable, and generic explanations, resulting in uninformative rationales. Further, our analysis shows that previous models' generated content often contain factual hallucinations. These issues call for novel solutions that could generate both informative and factually grounded explanations. Inspired by recent success in using retrieved content in addition to parametric knowledge for generation, we propose to augment the generator with a personalized retriever, where the retriever's output serves as external knowledge for enhancing the generator. Experiments on Yelp, TripAdvisor, and Amazon Movie Reviews dataset show our model could generate explanations that more reliably entail existing reviews, are more diverse, and are rated more informative by human evaluators.

MarioGPT OpenEnded Text2Level Generation through Large Language Models ; Procedural Content Generation PCG algorithms provide a technique to generate complex and diverse environments in an automated way. However, while generating content with PCG methods is often straightforward, generating meaningful content that reflects specific intentions and constraints remains challenging. Furthermore, many PCG algorithms lack the ability to generate content in an openended manner. Recently, Large Language Models LLMs have shown to be incredibly effective in many diverse domains. These trained LLMs can be finetuned, reusing information and accelerating training for new tasks. In this work, we introduce MarioGPT, a finetuned GPT2 model trained to generate tilebased game levels, in our case Super Mario Bros levels. We show that MarioGPT can not only generate diverse levels, but can be textprompted for controllable level generation, addressing one of the key challenges of current PCG techniques. As far as we know, MarioGPT is the first texttolevel model. We also combine MarioGPT with novelty search, enabling it to generate diverse levels with varying playstyle dynamics i.e. player paths. This combination allows for the openended generation of an increasingly diverse range of content.