Hao Zhang

Diffusion LLMs Can Do Faster-Than-AR Inference via Discrete Diffusion Forcing

Deep Think with Confidence

lmgame-Bench: How Good are LLMs at Playing Games?

Done Is Better than Perfect: Unlocking Efficient Reasoning by Structured Multi-Turn Decomposition

Stronger-MAS: Multi-Agent Reinforcement Learning for Collaborative LLMs

AR Models can be Faster and More Accurate Parallel Decoders than Diffusion LLMs

XYZ-Text2SQL-R1: Simple Rewards, Strong Reasoning in Text-to-SQL

Faster Video Diffusion with Trainable Sparse Attention

Efficiently Scaling LLM Reasoning Programs with Certaindex

Fast Video Generation with Sliding Tile Attention

GameArena: Evaluating LLM Reasoning through Live Computer Games

MPC-Minimized Secure LLM Inference

MatryoshkaKV: Adaptive KV Compression via Trainable Orthogonal Projection

Scaling Speculative Decoding with Lookahead Reasoning

Efficient-vDiT: Efficient Video Diffusion Transformers With Attention Tile

Scaling Long Context Training Data by Long-Distance Referrals

Cost-Effective Synthetic Data Generation for Post-Training using QWICK

In-context KV-Cache Eviction for LLMs via Attention-Gate