Liang He

Dynamic Multimodal Activation Steering for Hallucination Mitigation in Large Vision-Language Models

InternAgent-MLE: Navigating Fine-Grained Optimization for Coding Agent

Code-driven Number Sequence Calculation: Enhancing the Inductive Reasoning Abilities of Large Language Models

From Traits to Circuits: Toward Mechanistic Interpretability of Personality in Large Language Models

A Benchmark for Self-Evolving Agents via Experience-Driven Lifelong Learning

Protein-SE(3): Unified Framework and Comprehensive Benchmark for SE(3)-based Protein Structure Design

Cross-modal Alignment and Human Preference Learning for Fine-grained Music-guided Image Generation

A2A: Mechanistic Analysis for Efficient Layer Selection in Activation Steering

DualResearch: Entropy-Gated Dual-Graph Retrieval for Answer Reconstruction

Accelerating Diffusion Model with Dynamic Alignment

Reasoning Like Humans: Enhancing Multi-Image Reasoning Via Cognition-Inspired Meta-Action Framework

DriveArena: A Closed-loop Generative Simulation Platform for Autonomous Driving

Human Simulacra: Benchmarking the Personification of Large Language Models