Hugging Face's logo
Bill Yuchen Lin's picture
50 24 146

Bill Yuchen Lin

yuchenlin
yuntian-deng's profile picture r-ita's profile picture Agnuxo's profile picture
·
https://yuchenlin.xyz

AI & ML interests

Research @allenai LLMs and Multimodality, Agents

Articles

Organizations

Ai2's profile picture Multimodal Art Projection's profile picture INK Lab @ USC's profile picture Blog-explorers's profile picture LLM Blender's profile picture LoraHub's profile picture Human_Eval_RLHF's profile picture Lumos Agents (AI2)'s profile picture Qwen's profile picture ZeroGPU Explorers's profile picture SwiftSage Agent's profile picture Re-Align Project @ AI2's profile picture Agent-ETO's profile picture OLMo Friends's profile picture SimulBench's profile picture WildEval's profile picture Commonsense Index Team @ AI2's profile picture WildVision Team's profile picture Re-Eval-MMLU's profile picture AI2 Adapt Dev's profile picture Magpie Alignment's profile picture Social Post Explorers's profile picture Chinese LLMs on Hugging Face's profile picture Magpie's profile picture

Posts 2

view post
Post
2279
Introducing **BaseChat**!
https://huggingface.co/spaces/allenai/BaseChat_URIAL

This is a demo for our URIAL paper that enables base LLMs to chat with in-context alignment. You can talk directly with base, untuned LLMs to find out what knowledge and skills they have already learned from pre-training instead of SFT or xPO or RLHF. Also, you can use this to explore the pre-training data of base LLMs by chatting! I found a very interesting case: Base version of Llama-3-8B often thinks it is built by OpenAI, lol.
view post
Post
Introducing Vision Arena (beta)! Based on the lmsys's ChatbotArena, we create a simple demo for testing different Vision LMs (VLMs). We now support GPT-4V, Gemini-Pro-Vision, and Llava. More updates and models will come soon! We are still in the development stage and for now and we'd love to hear your feedback and suggestions! Please help us vote for better VLMs in your own use cases here! :D Kudos to Yujie Lu (UCSB)!
WildVision/vision-arena

Papers 21