arXiv AI

What Matters for Aggressive Decoding-Time KV Eviction? Temporal Aggregation and Ranking Preservation

The paper investigates how temporal aggregation and ranking preservation affect aggressive decoding-time KV cache eviction. It shows that exponential‑moving‑average (EMA) aggregation makes many scorer modifications indistinguishable at the eviction‑set level, while certain variants (KeyDiff, key norm, recency, learned scorer) alter rankings and degrade performance. Building on this, the authors propose InertiaKV and its lazy variant, achieving 1.34–1.46× faster decoding, and also explore a score‑free decoding approach that slightly improves quality while eliminating further scoring.

Hugging Face Trending Papers
Sep 3

What Matters for Aggressive Decoding-Time KV Eviction? Temporal Aggregation and Ranking Preservation

The paper investigates how temporal aggregation of token scores during decoding-time KV cache compression affects eviction decisions. It shows that using an exponential‑moving‑average (EMA) aggregation keeps ranking changes minimal for many scorer variants, while others like KeyDiff or recency significantly alter rankings and degrade performance. Building on this, the authors propose InertiaKV and its lazy variant, which achieve 1.34–1.46× faster decoding than full refresh, and also evaluate a score‑free approach that slightly improves quality while eliminating further scoring.

arXiv Computation and Language
Aug 28

NestedKV: Nested Memory Routing for Long-Context KV Cache Compression

NestedKV is a training‑free key‑only KV cache compression technique for long‑context language models that uses global, block‑level, and sliding‑window key anchors to score tokens via multi‑time‑scale cosine anomaly. It combines these rankings with a head‑adaptive outer learner and surprise‑gated token routing, requiring no model modification or additional training. Experiments on Qwen3 and Llama‑3.2 across benchmarks such as RULER, LongBench, and MMLU‑Pro show that NestedKV outperforms existing methods when the retained cache is small, achieving up to 19‑point gains on RULER and LongBench at a 75% retention rate.

By Hong Chen, Xiang Liu, Yubo Gao, Yuxuan Fan, Bo Wang, Yuanlin Chu, Yuanguo Lin, Xuming Hu
arXiv Machine Learning
Sep 4

Beyond Endpoint Scores: Time- and Capacity-Conditioned Evaluation of Continual Knowledge Updating

The paper argues that evaluating continual knowledge‑updating methods solely at a final checkpoint and a single adapter rank can be misleading. By fixing a periodic hierarchy and comparing it to cumulative replay on a 24‑month Wikidata stream, the authors show that the apparent best method changes depending on the evaluation month, replay LoRA rank, and query formulation. They recommend reporting performance trajectories and capacity sweeps, and only declaring a robust winner when the ranking remains stable across the evaluation region.

By Heejin Choi
arXiv Machine Learning
Aug 27

Trust the Mass: Forced Weights in KV-Cache Eviction

The paper investigates KV‑cache eviction strategies for sparse‑attention models, showing that selecting the largest attention weights is nearly optimal—closing only a median 2–5 % of the gap to full attention. It further demonstrates that differences in performance between eviction methods largely stem from memory usage, with the new training‑free ContourKV allocator outperforming state‑of‑the‑art methods in most pairwise comparisons while matching their byte‑efficiency.

By Jack Shi, Jerry Gu
arXiv AI
Jul 7

IndexMem: Learned KV-Cache Eviction with Latent Memory for Long-Context LLM Inference

arXiv:2605. 25475v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly expected to operate over long contexts, yet standard softmax attention incurs a KV cache that grows linearly with sequence length, quickly becoming the bottleneck for long context inference.

By Xintong Yang, Hao Gu, Binxing Xu, Lujun Li, Bei Liu, Jiacheng Liu, Qiyuan Zhu, Yike Guo, Sirui Han