MMFlash: Faster Speculative Decoding for Vision-Language Models Paper • 2611.18482 • Published 6 days ago • 41
MMFlash Collection MMFlash draft models for multimodal speculative decoding (Qwen3.5-4B/9B, Gemma 4 E4B-it, Qwen3.8-27B) • 5 items • Updated 3 days ago • 9
SIEVE: Preserving Cross-Image Evidence via Depth-Aware Token Compression Paper • 2611.00351 • Published 20 days ago • 1
DynaFlow: Learning Free-Form Multi-Agent Orchestration Paper • 2610.34467 • Published 25 days ago • 41
CoVeR: Coverage-Based Token Pruning for Multi-View 3D Reasoning in VLMs Paper • 2609.08345 • Published Sep 8 • 10
PISCES: Annotation-free Text-to-Video Post-Training via Optimal Transport-Aligned Rewards Paper • 2602.01624 • Published Feb 2 • 22
What about gravity in video generation? Post-Training Newton's Laws with Verifiable Rewards Paper • 2512.00425 • Published Nov 29, 2025 • 52