Dongrui Liu

ExGRPO: Learning to Reason from Experience

Your Agent May Misevolve: Emergent Risks in Self-evolving LLM Agents

EVCtrl: Efficient Control Adapter for Visual Generation

Accelerating Diffusion Large Language Models with SlowFast Sampling: The Three Golden Principles

OrthAlign: Orthogonal Subspace Decomposition for Non-Interfering Multi-Objective Alignment

Towards Self-Evolving Agent Benchmarks : Validatable Agent Trajectory via Test-Time Exploration

Reasoning over Boundaries: Enhancing Specification Alignment via Test-time Delibration

The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMs

Conditional Advantage Estimation for Reinforcement Learning in Large Reasoning Models

Evaluating Data Influence in Meta Learning

Can LLMs Reason Soundly in Law? Auditing Inference Patterns for Legal Judgment

Rethinking Entropy Regularization in Large Reasoning Models

Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolution

PACEbench: A Framework for Evaluating Practical AI Cyber-Exploitation Capabilities

REEF: Representation Encoding Fingerprints for Large Language Models

RiOSWorld: Benchmarking the Risk of Multimodal Computer-Use Agents

Demystifying Reasoning Dynamics with Mutual Information: Thinking Tokens are Information Peaks in LLM Reasoning

Derail Yourself: Multi-turn LLM Jailbreak Attack through self-discovered clues

Multi-Frame Neural Scene Flow: Learning Bounds and Algorithms

A Tailored Framework for Aligning Diffusion Models with Human Preference

Decouple-Then-Merge: Towards Better Training for Diffusion Models

DEAN: Deactivating the Coupled Neurons to Mitigate Fairness-Privacy Conflicts in Large Language Models