Matthew Gwilliam

mgwillia's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted 20 papers 1 day ago

IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse

Paper • 2603.12201 • Published 4 days ago • 47

Believe Your Model: Distribution-Guided Confidence Calibration

Paper • 2603.03872 • Published 13 days ago • 38

Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs

Paper • 2603.09906 • Published 6 days ago • 64

Lost in Stories: Consistency Bugs in Long Story Generation by LLMs

Paper • 2603.05890 • Published 11 days ago • 83

Lost in Backpropagation: The LM Head is a Gradient Bottleneck

Paper • 2603.10145 • Published 6 days ago • 10

Are Video Reasoning Models Ready to Go Outside?

Paper • 2603.10652 • Published 6 days ago • 7

Mario: Multimodal Graph Reasoning with Large Language Models

Paper • 2603.05181 • Published 12 days ago • 8

EffectMaker: Unifying Reasoning and Generation for Customized Visual Effect Creation

Paper • 2603.06014 • Published 11 days ago • 9

PureCC: Pure Learning for Text-to-Image Concept Customization

Paper • 2603.07561 • Published 9 days ago • 9

From Narrow to Panoramic Vision: Attention-Guided Cold-Start Reshapes Multimodal Reasoning

Paper • 2603.03825 • Published 13 days ago • 10

RubiCap: Rubric-Guided Reinforcement Learning for Dense Image Captioning

Paper • 2603.09160 • Published 7 days ago • 12

Dynamic Chunking Diffusion Transformer

Paper • 2603.06351 • Published 11 days ago • 14

MM-Zero: Self-Evolving Multi-Model Vision Language Models From Zero Data

Paper • 2603.09206 • Published 7 days ago • 46

Stepping VLMs onto the Court: Benchmarking Spatial Intelligence in Sports

Paper • 2603.09896 • Published 7 days ago • 25

RoboMME: Benchmarking and Understanding Memory for Robotic Generalist Policies

Paper • 2603.04639 • Published 12 days ago • 24

Progressive Residual Warmup for Language Model Pretraining

Paper • 2603.05369 • Published 12 days ago • 33

Just-in-Time: Training-Free Spatial Acceleration for Diffusion Transformers

Paper • 2603.10744 • Published 6 days ago • 7

CoCo: Code as CoT for Text-to-Image Preview and Rare Concept Generation

Paper • 2603.08652 • Published 7 days ago • 33

Fish Audio S2 Technical Report

Paper • 2603.08823 • Published 7 days ago • 27

Geometric Autoencoder for Diffusion Models

Paper • 2603.10365 • Published 6 days ago • 6