Bill Yuchen Lin
agent training
llm
ensemble learning
open-source
decoding
langauge generation
llm agent
large language models
vision-language reasoning; selective prediction; confidence estimation
agent
4
presentations
SHORT BIO
Yuchen Lin is a Young Investigator at Allen Institute for AI (AI2), hosted by Prof. Yejin Choi (University of Washington). He received his PhD from University of Southern California in 2022, advised by Prof. Xiang Ren. Yuchen's primary interest lies in developing artificial intelligence agents capable of demonstrating a profound understanding of the world through common-sense knowledge and reasoning abilities. His research aims to teach machines how to think, talk, and act like humans. Moreover, Yuchen's work focuses on enhancing the robustness, safety, and generalization of large language models (LLMs) through retrieval augmentation, continual learning, federated learning, and ensemble learning techniques.
Presentations
Trial and Error: Exploration-Based Trajectory Optimization of LLM Agents
Yifan Song and 5 other authors
Selective “Selective Prediction”: Reducing Unnecessary Abstention in Vision-Language Reasoning
Tejas Srinivasan and 6 other authors
Agent Lumos: Unified and Modular Training for Open-Source Language Agents
Da Yin and 6 other authors
LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusion
Bill Yuchen Lin and 2 other authors