Huan Wang

Autoregressive Image Generation with Randomized Parallel Decoding

MergeMix: A Unified Augmentation Paradigm for Visual and Multi-Modal Understanding

RewardMap: Tackling Sparse Rewards in Fine-grained Visual Reasoning via Multi-Stage Reinforcement Learning

SparseSSM: Efficient Selective Structured State Space Models Can Be Pruned in One-Shot

Which Heads Matter for Reasoning? RL-Guided KV Cache Compression

OBS-Diff: Accurate Pruning For Diffusion Models in One-Shot

SSD: Spatial-Semantic Head Decoupling for Efficient Autoregressive Image Generation

MPP: MODEL POWER PARITY FOR REASONING LARGE LANGUAGE MODEL COMPARISON ON CROSS-DOMAIN BENCHMARKS

Boosting Large Language Models with Mask Fine-Tuning

ConCuR: Conciseness Makes State-of-the-Art Kernel Generation

Plug-and-Play 1.x-Bit KV Cache Quantization for Video Large Language Models

TARS: MinMax Token-Adaptive Preference Strategy for MLLM Hallucination Reduction

Can MLLMs Guide Me Home? A Benchmark Study on Fine-grained Visual Reasoning from Transit Maps

ARC-Decode: Risk-Bounded Acceptance for Sampling-Based Speculative Decoding

Reordered SparseGPT: Optimizing the Pruning Order in Second-Order LLM Pruning

TVMamba:Towards Efficient Visual Mammba With Ternary Weights and Activations

ERC-SVD: Error-Controlled SVD for Large Language Model Compression

FreqExit: Enabling Early-Exit Inference for Visual Autoregressive Models via Frequency-Aware Guidance

Accessing Vision Foundation Models via ImageNet-1K

HoliTom: Holistic Token Merging for Fast Video Large Language Models

Poison as Cure: Visual Noise for Mitigating Object Hallucinations in LVMs

Segment then Splat: Unified 3D Open-Vocabulary Segmentation via Gaussian Splatting