# KVarN: Variance-Normalized KV-Cache Quantization Mitigates Error Accumulation in Reasoning Tasks

Page: https://stenobird.com/podcast/daily-paper-cast-7079649/kvarn-variance-normalized-kv-cache-quantization-mitigates-error-accumulation-in-reasoning-tasks
Text version: https://stenobird.com/podcast/daily-paper-cast-7079649/kvarn-variance-normalized-kv-cache-quantization-mitigates-error-accumulation-in-reasoning-tasks.md
Podcast: [Daily Paper Cast](https://stenobird.com/podcast/daily-paper-cast-7079649)
Published: 2026-06-04T03:55:36+00:00
Episode link: https://share.transistor.fm/s/397d6a64
Audio file: https://media.transistor.fm/397d6a64/2055afea.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/kvarn-variance-normalized-kv-cache-quantization-mitigates-error-accumulation-in-reasoning-tasks
Duration seconds: 1360

## Resource

🤗 Upvotes: 27 | cs.LG Authors: Lorenz K. Muller, Philippe Bich, Chiara Boretti, Hyun-Min Chang, Jiawei Zhuang, Lukas Cavigelli Title: KVarN: Variance-Normalized KV-Cache Quantization Mitigates Error Accumulation in Reasoning Tasks Arxiv: http://arxiv.org/abs/2606.03458v1 Abstract: Test-time scaling is a powerful approach to obtain better reasoning in large language models, but it becomes memory-bottlenecked during long-horizon decoding, as the KV-cache grows. KV-cache quantization can help improve this, but current methods are evaluated under prefill-like settings and errors behave differently under autoregressive decoding. We show that in the latter regime, quantization errors accumulate across timesteps, driven primarily by incorrect token scales. We introduce KVarN, a calibration-free KV-cache quantizer that applies a Hadamard rotation followed by a dual-scaling variance normalization across both axes of the K and V matrices. We find that this combination fixes outlying token-scale errors and substantially reduces error accumulation over existing baselines. KVarN establishes a new state-of-theart for KV-cache quantization on generative benchmarks, including MATH500, AIME24 and HumanEval, at 2-bit precision. A vLLM implementation of the KVarN method is available at https://github.com/huawei-csl/KVarN

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/kvarn-variance-normalized-kv-cache-quantization-mitigates-error-accumulation-in-reasoning-tasks/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/daily-paper-cast-7079649/kvarn-variance-normalized-kv-cache-quantization-mitigates-error-accumulation-in-reasoning-tasks.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.