Mengdi Wang

FutureX: An Advanced Live Benchmark for LLM Agents in Future Prediction

ARMOR: High-Performance Semi-Structured Pruning via Adaptive Matrix Factorization

GeneBreaker: Jailbreak Attacks against DNA Language Models with Pathogenicity Guidance

Alita-G: Self-Evolving Generative Agent for Agent Generation

MMaDA-Parallel: Multimodal Large Diffusion Language Models for Thinking-Aware Editing and Generation

FMIP: Joint Continuous-Integer Flow For Mixed-Integer Linear Programming

CubeBench: Diagnosing Interactive, Long-Horizon Physical Intelligence under Partial Observations

Revolutionizing Reinforcement Learning Framework for Diffusion Large Language Models

PoseX: AI Defeats Physics-based Methods on Protein Ligand Cross-Docking

Teaching Language Model to Act Efficiently

AutoTool: Dynamic Tool Selection and Integration for Agentic Reasoning

Detecting Post-generation Edits to Watermarked LLM Outputs via Combinatorial Watermarking

SafeProtein: Red-Teaming Framework and Benchmark for Protein Foundation Models

On the Role of Preference Variance in Preference Optimization

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

Demystifying Reinforcement Learning in Agentic Reasoning

On Path to Multimodal Historical Reasoning: HistBench and HistAgent

SafeThink: A Key to Safety in Multi-Modal Large Reasoning Models

AgentDistill: Training-Free Agent Distillation with Generalizable MCP Boxes

CURE: Co-Evolving Coders and Unit Testers via Reinforcement Learning

Does Thinking More Always Help? Mirage of Test-Time Scaling in Reasoning Models

ReasonFlux-PRM: Trajectory-Aware PRMs for Long Chain-of-Thought Reasoning in LLMs

Training-Free Guidance Beyond Differentiability: Scalable Path Steering with Tree Search in Diffusion and Flow Models

DISC: Dynamic Decomposition Improves LLM Inference Scaling

MMaDA: Multimodal Large Diffusion Language Models

IterComp: Iterative Composition-Aware Feedback Learning from Model Gallery for Text-to-Image Generation

Securing the Language of Life: Inheritable Watermarks from DNA Language Models to Proteins

Diffusion Transformer Captures Spatial-Temporal Dependencies: A Theory for Gaussian Process Data

NoWag: A Unified Framework for Shape Preserving Com- pression of Large Language Models

SpecDec++: Boosting Speculative Decoding via Adaptive Candidate Lengths

Emergent Symbolic Mechanisms Support Abstract Reasoning in Large Language Models

A First-order Generative Bilevel Optimization Framework for Diffusion Models

Collab: Controlled Decoding using Mixture of Agents for LLM Alignment

MATH-Perturb: Benchmarking LLMs' Math Reasoning Abilities against Hard Perturbations

A Common Pitfall of Margin-based Language Model Alignment: Gradient Entanglement

Towards Understanding Text Hallucination of Diffusion Models via Local Generation Bias

Rectified Diffusion: Straightness Is Not Your Need in Rectified Flow

SAIL: Self-improving Efficient Online Alignment of Large Language Models

SpecDec++: Boosting Speculative Decoding via Adaptive Candidate Lengths

Online Detection for Black-Box Large Language Models with Adaptive Prompt Selection

Regularized DeepIV with Model Selection

LIAR: Leveraging Inverse Alignment to Jailbreak LLMs in Seconds

Relative-Translation Invariant Wasserstein Distance

AIME: AI System Optimization via Multiple LLM Evaluators

GuideCO: Training Objective-Guided Diffusion Solver with Imperfect Data for Combinatorial Optimization