Paul Pu Liang

World-In-World: World Models in a Closed-Loop World

Building Massively Multimodal Foundation Models with Interaction-aware Mixture-of-Experts

MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon Agents

PAGE-4D: Disentangled Pose and Geometry Estimation for VGGT-4D Perception

RAVENEA: A Benchmark for Multimodal Retrieval-Augmented Visual Culture Understanding

PuzzleWorld: A Benchmark for Multimodal, Open-Ended Reasoning in Puzzlehunts

Human Behavior Atlas: Benchmarking Unified Psychological And Social Behavior Understanding

FairGRPO: Fair Reinforcement Learning for Equitable Clinical Reasoning

SmellNet: A Dataset for Sensor-Based Smell Recognition and Mixture Prediction

When Reasoning Meets Its Laws

MENTALMs: Teaching Language Models to Reason Efficiently in the Language of Thought

Learn Globally, Speak Locally: Bridging the Gaps in Multilingual Reasoning

HugAgent: Evaluating LLMs in Simulating Human-Like Individual Reasoning on Open-Ended Tasks

CTM-AI: A Blueprint for General AI Inspired by Consciousness

ChartRef: Benchmarking Fine-Grained Visual Element Localization in Charts

EmoSign: A Multimodal Dataset for Understanding Emotions in American Sign Language

MULTI-FACETED MULTIMODAL MONOSEMANTICITY

Sotopia-RL: Reward Design for Social Intelligence

QoQ-Med: Building Multimodal Clinical Foundation Models with Domain-Aware GRPO Training

Balancing Multimodal Training Through Game-Theoretic Regularization

Progressive Compositionality in Text-to-Image Generative Models

OS-ATLAS: Foundation Action Model for Generalist GUI Agents

Can A Society of Generative Agents Simulate Human Behavior and Inform Public Health Policy? A Case Study on Vaccine Hesitancy

Partial Information Decomposition via Normalizing Flows in Latent Gaussian Distributions

What One Cannot, Two Can: Two-Layer Transformers Provably Represent Induction Heads on Any-Order Markov Chains

REGen: Multimodal Retrieval-Embedded Generation for Long-to-Short Video Editing

VideoWebArena: Evaluating Long Context Multimodal Agents with Video Understanding Web Tasks

Understanding the Emergence of Multimodal Representation Alignment

CLIMB: Data Foundations for Large Scale Multimodal Clinical Foundation Models

Video Active Perception: Efficient Inference-Time Long-Form Video Understanding with Vision-Language Models

TeaserGen: Generating Teasers for Long Documentaries