Jianfeng Gao

FlowRL: Matching Reward Distributions for LLM Reasoning

Dyna-Mind: Learning to Simulate from Experience for Better AI Agents

Training Large Reasoning Models Efficiently via Progressive Thought Encoding

Dyna-Think: Synergizing Reasoning, Acting, and World Model Simulation in AI Agents

SEMA: Simple yet Effective Learning for Multi-Turn Jailbreak Attacks

WidgetEval: Benchmarking Foundation Models on Dynamic Widget Generation for Apps

DenseMixer: Improving MoE Post-Training with Precise Router Gradient

TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models

MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Stress-testing LLM Safety

EfficientLLM: Evaluating Large Language Models Efficiency

TemporalBench: Evaluating Fine-Grained Temporal Dynamics Understanding for Multimodal Models

Coupling Attention and Memory: A Dynamic Memory Module for Efficient Adapation with Pretrained LLMs

Mixture of Inputs: Text Generation Beyond Discrete Token Sampling

CollabLLM: From Passive Responders to Active Collaborators

Interpretable Next-token Prediction via the Generalized Induction Head

Reinforcement Learning for Reasoning in Large Language Models with One Training Example

GUI-Actor: Coordinate-Free Visual Grounding for GUI Agents

TraceVLA: Visual Trace Prompting Enhances Spatial-Temporal Awareness for Generalist Robotic Policies

MMInference: Accelerating Pre-filling for Long-Context Visual Language Models via Modality-Aware Permutation Sparse Attention

Decoder-Hybrid-Decoder Architecture for Efficient Reasoning with Long Generation

Interpretable Language Modeling via Induction-head Ngram Models

SCBench: A KV Cache-Centric Analysis of Long-Context Methods

Elevating Visual Perception in Multimodal LLMs with Visual Embedding Distillation

Training Language Models to Generate Quality Code with Program Analysis Feedback

SAS: Simulated Attention Score

GUI-World: A Video Benchmark and Dataset for Multimodal GUI-oriented Understanding

Simplifying DINO via Coding Rate Regularization

Matryoshka Multimodal Models

Vector-ICL: In-context Learning with Continuous Vector Representations

DataGen: Unified Synthetic Dataset Generation via Large Language Models

Latent Action Pretraining from Videos

ExACT: Teaching AI Agents to Explore with Reflective-MCTS and Exploratory Learning

Generative Adapter: Contextualizing Language Models in Parameters with A Single Forward Pass

SeCom: On Memory Construction and Retrieval for Personalized Conversational Agents

Model Tells Itself Where to Attend: Steerable Prompting for Reliable Reading Comprehension of LLM

Riemannian Low-Rank Adaptation for Federated Fine-Tuning of Foundation Models

TemporalBench: Towards Fine-grained Temporal Understanding for Multimodal Video Models

Pixelated Instructions: Can Multimodal Large Language Models Follow Printed Instructions in Images?

Evaluating Graphical Perception of Large Multimodal Models