prompt-compression

Tag

Cards List
#prompt-compression

Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression

arXiv cs.CL · 2026-08-06 Cached

This paper identifies a structural failure in hard prompt compression called 'referential dangling', where independent scoring splits dependent evidence pairs, removing the context needed to interpret retained answers. Experiments show this affects many compressors and datasets, and automatic restoration of missing references improves QA accuracy.

0 favorites 0 likes
#prompt-compression

What Is a Skill Worth? Structure-Aware Shapley Valuation of Agent Skills

arXiv cs.AI · 2026-08-06 Cached

This paper introduces SkillSV, a structure-aware Shapley-style framework for valuing the internal units of agent skills, accounting for dependencies and hierarchy. It demonstrates improved faithfulness, actionability, and explanation quality on agentic benchmarks.

0 favorites 0 likes
#prompt-compression

Every Time I Hire a Linguist, Inference Costs Go Down: On Linguistic Rules as Effective Prompt Compressors

arXiv cs.CL · 2026-07-29 Cached

This paper proposes using linguistic rules alone as prompt compressors without LM forward passes, achieving performance similar to advanced strategies under light-to-moderate compression.

0 favorites 0 likes
#prompt-compression

Cache-Aware Prompt Compression:A Two-Tier Cost Model for LLM API Caching

arXiv cs.AI · 2026-07-20 Cached

Proposes Cache-Aware Prompt Compression (CAPC), a method that combines query-agnostic compression with caching to reduce LLM API costs, demonstrating significant savings over existing approaches on Anthropic's Sonnet API and production workloads.

0 favorites 0 likes
#prompt-compression

What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents

arXiv cs.LG · 2026-07-10 Cached

This paper unifies memory compaction techniques across LLMs and agents under a rate-distortion framework, proposing a taxonomy and benchmark for evaluating compression across different layers.

0 favorites 0 likes
#prompt-compression

Prompt Compression via Activation Aggregation

arXiv cs.CL · 2026-07-10 Cached

This paper proposes compressing instruction prompts into a single activation vector via learned weighted sums of intermediate layer activations, achieving under 2% accuracy drop and revealing insights into LLM activation space structure.

0 favorites 0 likes
#prompt-compression

Mapping Text to Multiplex Graph: Prompt Compression as L\'evy Walk-Guided Graph Pruning

arXiv cs.CL · 2026-07-03 Cached

This paper proposes RAGP, a prompt compression method that models text as a multiplex graph and uses Lévy walks for redundancy-aware graph pruning, achieving superior performance on LongBench compared to existing LLM-based compression methods.

0 favorites 0 likes
#prompt-compression

Probing the Prompt KV Cache: Where It Becomes Dispensable

arXiv cs.CL · 2026-06-01 Cached

This paper systematically investigates when and which parts of the prompt KV cache become dispensable during LLM decoding, showing that redundancy primarily involves chat template scaffolding rather than task content, and replacement with neutral filler preserves accuracy.

0 favorites 0 likes
#prompt-compression

AGORA: Adapter-Grounded Observation-Action Retention for Inference-Free Prompt Compression in LLM Agents

arXiv cs.AI · 2026-05-27 Cached

AGORA introduces an inference-free step-level prompt compressor for LLM agents that avoids the 'action-grammar destruction' failure mode of token-level compressors. It retains ≥75% uncompressed performance in 8 of 9 environments across backbones, using a structural parser, an always-keep floor, and a learned relevance scorer.

0 favorites 0 likes
#prompt-compression

From Token to Token Pair: Efficient Prompt Compression for Large Language Models in Clinical Prediction

arXiv cs.CL · 2026-05-13 Cached

This paper introduces MedTPE, a method for efficient, lossless prompt compression of electronic health records for large language models, significantly reducing token length and inference latency in clinical prediction tasks.

0 favorites 0 likes
← Back to home

Submit Feedback