Papers with prefilling and decoding

2 papers
SpecVLM: Enhancing Speculative Decoding of Video LLMs via Verifier-Guided Token Pruning (2025.emnlp-main)

Copied to clipboard

Challenge: Video large language models (Vid-LLMs) rely on dense video token representations and require substantial memory and computational overhead in both prefilling and decoding.
Approach: They propose a training-free speculative decoding framework that prunes up to 90% of video tokens to enable efficient speculation without sacrificing accuracy.
Outcome: The proposed framework achieves 2.68 speedup on LLaVA-OneVision-72B and 2.11 speed up on Qwen2.5-VL-32B.
Accelerating Adaptive Retrieval Augmented Generation via Instruction-Driven Representation Reduction of Retrieval Overlaps (2025.findings-acl)

Copied to clipboard

Challenge: Existing approaches to augmented generation ignore the overlap in retrieval results . overlapping content is redundantly represented, affecting the overall efficiency.
Approach: They propose a model-agnostic approach to re-augmented generation that speeds up prefilling and decoding . they propose an instruction-driven module to guide the model to more suitable ways for LLMs .
Outcome: The proposed approach achieves 2.79 and 2.33 times significant acceleration on average for prefilling and decoding respectively while maintaining equal generation quality.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations