analiza aiqbench
The paper proposes EffComp, a sentence-level extractive method for compressing prompts in retrieval-augmented question answering, using BERT-style reranker and selector models trained with a mix of reinforcement and supervised learning. Its main claim is a better compression-accuracy trade-off and lower latency than existing methods, but these results are reported on the authors' own QA benchmarks and need independent replication. For practitioners, the useful point is that extractive compression avoids generating new text, which limits hallucination risk in the compressed prompt.
EffComp: Efficient Prompt Compression via Hybrid Reinforcement & Supervised Learning
arXiv:2611.02766v1 Announce Type: new Abstract: Prompt compression is aimed at reducing input prompt lengths to enable cheaper, faster, and potentially higher-quality LLM predictions. However, existing prompt compression methods are often limited by modest compression gains, a risk of hallucination, and/or high compression latency. We propose EffComp, an Efficient prompt Compression framework for RAG-based question answering (QA). EffComp employs BERT-style document reranker and sentence selector models to allow fast extractive prompt compression at the sentence level. Its extractive nature prevents hallucinations in the compressed prompts. Additionally, the training process using a hybrid reinforcement and supervised learning approach is designed to optimize the compression ratio while preserving LLM accuracy. Experiments on various QA datasets demonstrate that EffComp outperforms state-of-the-art prompt compression methods achieving the best trade-off between compression ratio and accuracy with minimal latency. This makes it practical for real-world applications.
Abstract:Prompt compression is aimed at reducing input prompt lengths to enable cheaper, faster, and potentially higher-quality LLM predictions. However, existing prompt compression methods are often limited by modest compression gains, a risk of hallucination, and/or high compression latency. We propose EffComp, an Efficient prompt Compression framework for RAG-based question answering (QA). EffComp employs BERT-style document reranker and sentence selector models to allow fast extractive prompt compression at the sentence level. Its extractive nature prevents hallucinations in the compressed prompts. Additionally, the training process using a hybrid reinforcement and supervised learning approach is designed to optimize the compression ratio while preserving LLM accuracy. Experiments on various QA datasets demonstrate that EffComp outperforms state-of-the-art prompt compression methods achieving the best trade-off between compression ratio and accuracy with minimal latency. This makes it practical for real-world applications.
Computation and Language (cs.CL)
arXiv:2611.02766 [cs.CL]
(or arXiv:2611.02766v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2611.02766 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 229 cuvinte
Subiecte conexe
Think, Compress, Answer: Reinforcement Learning for Multimodal Grounding via Reasoning Compression
arXiv · 20 oct. 2026
Reward function compression facilitates goal-dependent reinforcement learning
Nature · 8 oct. 2026
Explainable Machine Learning Analysis of Temperature Effects on Mechanical Strength of Fiber-Reinforced High-Performance Concrete
arXiv · 19 oct. 2026
Agent Skill Evaluation and Evolution: Frameworks and Benchmarks
arXiv · 3 nov. 2026