Bowen Jin

GRACE: Generative Representation Learning via Contrastive Policy Optimization

Demystifying and Enhancing the Efficiency of Large Language Model Based Search Agents

mCLM: A Modular Chemical Language Model that Generates Functional and Makeable Molecules

Learning to Reason as Action Abstractions with Scalable Mid-Training RL

RM-R1: Reward Modeling as Reasoning

Teaching Language Model to Act Efficiently

COMPASS: A Multi-Turn Benchmark for Tool-Mediated Planning & Preference Optimization

Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning

Hybrid Latent Reasoning via Reinforcement Learning

LLM Alignment as Retriever Optimization: An Information Retrieval Perspective

Long-Context LLMs Meet RAG: Overcoming Challenges for Long Inputs in RAG

Benchmarking Retrieval-Augmented Generation for Chemistry

Multi-modal Prompt Learning Empowers Graph Neural Networks with Semantic Knowledge

CoDiCast: Conditional Diffusion Model for Weather Prediction with Uncertainty Quantification