Lei Li

Associate Professor@School of Computer Science, Carnegie Mellon University·美国·OpenReview

研究方向

LLM agent · molecule learning · protein design · drug discovery · large language model · machine translation · speech translation · machine learning. Bayesian methods · deep learning

Lei Li

JanusCoder: Towards a Foundational Visual-Programmatic Interface for Code Intelligence

HARDTESTGEN: A High-Quality RL Verifier Generation Pipeline for LLM Algorithimic Coding

RECAP: Reproducing Copyrighted Data from LLMs Training with an Agentic Pipeline

Can Large Language Models Assess and Reframe Psychological Attribution: A Benchmark and Analysis

Learning to Interrupt in Language-based Multi-agent Communication

RARE: Retrieval-Aware Robustness Evaluation for Retrieval-Augmented Generation Systems

Strategic Planning and Rationalizing on Trees Make LLMs Better Debaters

Is Vibe Coding Safe? Benchmarking Vulnerability of Agent Generated Code in Real-World Tasks

InteractScience: Programmatic and Visually-Grounded Evaluation of Interactive Scientific Demonstration Code Generation

TusoAI: Agentic Optimization for Scientific Methods

InstructPro: Natural Language Guided Ligand-Binding Protein Design

Permute-and-Flip: An optimally stable and watermarkable decoder for LLMs

Weak-to-Strong Jailbreaking on Large Language Models

Diversity Empowers Intelligence: Integrating Expertise of Software Engineering Agents

TypedThinker: Diversify Large Language Model Reasoning with Typed Thinking

Speculative Knowledge Distillation: Bridging the Teacher-Student Gap Through Interleaved Sampling

DNALONGBENCH: A Benchmark Suite For Long-Range DNA Prediction Tasks

PPDiff: Diffusing in Hybrid Sequence-Structure Space for Protein-Protein Complex Design

Weak-to-Strong Jailbreaking on Large Language Models

DIS-CO: Discovering Copyrighted Content in VLMs Training Data

Efficiently Identifying Watermarked Segments in Mixed-Source Texts