arXiv AI September 17, 2026 By Mingyang Mao, Wyatt Mackey, Xiaomin Lin

Contiguity, Not Importance: Budgeted Repair of Stale KV Caches After Document Edits

The paper investigates how to efficiently repair stale key-value (KV) caches in retrieval‑augmented generation systems after document edits. It proposes a budgeted in‑place recomputation approach and evaluates training‑free position‑selection policies on a factual RAG benchmark. Across three model families, a contiguous edit‑local window consistently recovers most of the post‑edit answer quality while being 13–21 times faster than a full re‑prefill, though its effectiveness diminishes when answer‑bearing text moves downstream.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv AI

Jun 17

Models Take Notes at Prefill: KV Cache Can Be Editable and Composable

arXiv:2606. 17107v1 Announce Type: cross Abstract: Prefix caching reuses prefill only across an exactly shared prefix, so one changed field invalidates the entire downstream cache.

By Bojie Li

agents efficiency multimodal benchmarks

arXiv Machine Learning

Jun 16

KVEraser: Learning to Steer KV Cache for Efficient Localized Context Erasing

arXiv:2606. 17034v1 Announce Type: cross Abstract: Post-hoc context erasing over the KV cache is challenging because a local edit has a global consequence: once a span has been processed, its influence propagates into the cached states of all subsequent tokens.

By Mufei Li, Shikun Liu, Dongqi Fu, Haoyu Wang, Yinglong Xia, Hong Li, Hong Yan, Pan Li

llms fine-tuning efficiency

arXiv Computation and Language

Sep 10

KVShareArena: KV-Cache Reuse Across Contexts and Model Checkpoints

arXiv:2609.10266v1 Announce Type: new Abstract: LLM serving systems already reuse KV caches, but only when the reused text sits at the very start of the prompt. Two growing workloads break this condi...

By Xi Shi, Qian Lou

llms rag agents benchmarks

arXiv Machine Learning

Aug 7

QEvict: Recoverable Quantized KV Eviction for Attention-Drift-Robust Long-Context Decoding

arXiv:2608. 05326v1 Announce Type: new Abstract: Autoregressive large language model inference is increasingly constrained by the memory footprint of the Key-Value (KV) cache.

By Ayushman Garg, Akshita Gupta, Shaswata Bhattacharya, Abhishek Gupta, Sandeep Kumar, Manoj Kumar

llms efficiency benchmarks

arXiv AI

Jun 30

Can Fine-Tuning Erase Your Edits? On the Fragile Coexistence of Knowledge Editing and Adaptation

arXiv:2511. 05852v4 Announce Type: replace-cross Abstract: Knowledge editing (KE) offers a lightweight alternative to retraining for updating large language models (LLMs).

By Yinjie Cheng, Paul Youssef, Christin Seifert, J\"org Schl\"otterer, Zhixue Zhao

llms fine-tuning benchmarks safety

arXiv AI

Sep 10

What Eviction Destroys: A Restore-Counterfactual Audit of Forgetting in Agent Memory

arXiv:2609.08279v1 Announce Type: cross Abstract: Agent memory systems must discard stored information when their history exceeds a fixed token budget. Existing budget-accuracy frontiers quantify the...

By Chen Shen

llms agents benchmarks

One email a morning, machine-written

Related stories

Models Take Notes at Prefill: KV Cache Can Be Editable and Composable

KVEraser: Learning to Steer KV Cache for Efficient Localized Context Erasing

KVShareArena: KV-Cache Reuse Across Contexts and Model Checkpoints

QEvict: Recoverable Quantized KV Eviction for Attention-Drift-Robust Long-Context Decoding

Can Fine-Tuning Erase Your Edits? On the Fragile Coexistence of Knowledge Editing and Adaptation

What Eviction Destroys: A Restore-Counterfactual Audit of Forgetting in Agent Memory