R2R: Efficiently Navigating Divergent Reasoning Paths with Small-Large Model Token Routing
Paper
• 2505.21600 • Published
• 71
Papers from the NICS-EFFALG Team
-
-
Distilled Decoding 1: One-step Sampling of Image Auto-regressive Models with Flow Matching
Paper • 2412.17153 • Published • 39 -
Skeleton-of-Thought: Large Language Models Can Do Parallel Decoding
Paper • 2307.15337 • Published • 39 -
DiTFastAttn: Attention Compression for Diffusion Transformer Models
Paper • 2406.08552 • Published • 25 -
Can LLMs Learn by Teaching? A Preliminary Study
Paper • 2406.14629 • Published • 21 -
MoA: Mixture of Sparse Attention for Automatic Large Language Model Compression
Paper • 2406.14909 • Published • 16 -
A Survey on Efficient Inference for Large Language Models
Paper • 2404.14294 • Published • 4 -
ViDiT-Q: Efficient and Accurate Quantization of Diffusion Transformers for Image and Video Generation
Paper • 2406.02540 • Published • 3 -
MixDQ: Memory-Efficient Few-Step Text-to-Image Diffusion Models with Metric-Decoupled Mixed Precision Quantization
Paper • 2405.17873 • Published • 3 -
FrameFusion: Combining Similarity and Importance for Video Token Reduction on Large Visual Language Models
Paper • 2501.01986 • Published • 1 -
Evaluating Quantized Large Language Models
Paper • 2402.18158 • Published -
Cache-to-Cache: Direct Semantic Communication Between Large Language Models
Paper • 2510.03215 • Published • 98