IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse
Paper • 2603.12201 • Published • 51
Matthew Gwilliam
mgwillia
AI & ML interests
Computer Vision
Recent Activity
upvoted a paper 7 days ago
IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reuse upvoted a paper 7 days ago
Believe Your Model: Distribution-Guided Confidence Calibration upvoted a paper 7 days ago
Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMsOrganizations
None yet
upvoted 20
papers 7 days ago
Believe Your Model: Distribution-Guided Confidence Calibration
Paper • 2603.03872 • Published • 39
Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs
Paper • 2603.09906 • Published • 71
Lost in Stories: Consistency Bugs in Long Story Generation by LLMs
Paper • 2603.05890 • Published • 91
Lost in Backpropagation: The LM Head is a Gradient Bottleneck
Paper • 2603.10145 • Published • 11
Are Video Reasoning Models Ready to Go Outside?
Paper • 2603.10652 • Published • 7
Mario: Multimodal Graph Reasoning with Large Language Models
Paper • 2603.05181 • Published • 8
EffectMaker: Unifying Reasoning and Generation for Customized Visual Effect Creation
Paper • 2603.06014 • Published • 9
PureCC: Pure Learning for Text-to-Image Concept Customization
Paper • 2603.07561 • Published • 9
From Narrow to Panoramic Vision: Attention-Guided Cold-Start Reshapes Multimodal Reasoning
Paper • 2603.03825 • Published • 10
RubiCap: Rubric-Guided Reinforcement Learning for Dense Image Captioning
Paper • 2603.09160 • Published • 15
Dynamic Chunking Diffusion Transformer
Paper • 2603.06351 • Published • 14
MM-Zero: Self-Evolving Multi-Model Vision Language Models From Zero Data
Paper • 2603.09206 • Published • 51
Stepping VLMs onto the Court: Benchmarking Spatial Intelligence in Sports
Paper • 2603.09896 • Published • 26
RoboMME: Benchmarking and Understanding Memory for Robotic Generalist Policies
Paper • 2603.04639 • Published • 28
Progressive Residual Warmup for Language Model Pretraining
Paper • 2603.05369 • Published • 36
Just-in-Time: Training-Free Spatial Acceleration for Diffusion Transformers
Paper • 2603.10744 • Published • 7
CoCo: Code as CoT for Text-to-Image Preview and Rare Concept Generation
Paper • 2603.08652 • Published • 39
Fish Audio S2 Technical Report
Paper • 2603.08823 • Published • 34
Geometric Autoencoder for Diffusion Models
Paper • 2603.10365 • Published • 6