AI ethics evidence index Search
Source-linked daily record
Evidence record 3316 · automatically gathered

Alignment Collapse Under KV Cache Quantization: Diagnosis and Mitigation

Key-value (KV) cache quantization is widely used to reduce Large Language Model (LLM) inference memory, yet existing evaluations solely focus on measuring perplexity and accuracy without assessing the safety impact. In this study, we explore alignment preservation under KV cache quantization. Across eleven instruction-tuned models (3.8B-72B) and five benchmarks (1,894 prompts), we find that low-bit quantization can silently destroy safety alignment: Mistral-7B loses 15.2% of its refusals at only

Record details

Published: 1 June 2026
Source: arXiv
Category: Research
Topics: Safety & alignment · Healthcare
Retrieved: 14 July 2026

source-onlyevidence status

These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.

Also mentions Mistral · Shared topic: Safety & alignment

Two Wrongs, No Right: Auditing Social-Desirability Bias in LLM Annotators for Computational Social Science

arXiv · 12 May 2026

Also mentions Mistral · Shared topic: Safety & alignment

Phoenix-VL 1.5 Medium Technical Report

arXiv · 11 May 2026

Also mentions Mistral · Shared topic: Healthcare

EQUITRIAGE: A Fairness Audit of Gender Bias in LLM-Based Emergency Department Triage

arXiv · 5 May 2026

Also mentions Mistral · Shared topic: Safety & alignment

Breaking Safety at the Token Boundary: How BPE Tokenization Creates Exploitable Gaps in LLM Alignment

arXiv · 1 May 2026

Also mentions Mistral · Shared topic: Safety & alignment

For What Reason? Interpreting Models' Encoding of Causation and Antithesis

arXiv · 20 July 2026

Also mentions Mistral · Shared topic: Healthcare

Do language families matter? Evaluating LLMs for sentiment analysis through a hierarchical cross-lingual lens

Frontiers in Artificial Intelligence · 24 July 2026

How to cite this record

ethics.ai (1 June 2026), “Alignment Collapse Under KV Cache Quantization: Diagnosis and Mitigation,” evidence record 3316, https://ethics.ai/record/3316 (originally published by arXiv).

JSON

Use and limitations

This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.