AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

The Heterogeneous Safety Impacts of Benign Multilingual Fine-Tuning

Will Hawkins, Kai Rawal, Jonathan Rystrøm, Stratis Tsirtsis, Zihao Fu, Greta Warren, Ryan Brown, Eoin D. Delaney, Sandra Wachter, Brent Mittelstadt, Chris Russell

ICML 2026 · Large language models › Instruction and fine-tuning · approach: Safety alignment after fine-tuning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
fine-tune on benign English data (alpha/r=8, 1 epoch)Gemma-3-1B-IT · SORRY-Bench English evalcompliance rate (%): 30.2 -> 65.5Hurt
fine-tune on benign Hindi data (alpha/r=8)Gemma-3-1B-IT · SORRY-Bench English evalcompliance rate (%): 44.8 -> 76.8Hurt
fine-tune on benign Portuguese dataGemma-3-4B-IT · SORRY-Bench local evalcompliance rate: about 4x increaseHurt
fine-tune on benign data (any language)Qwen3-0.6B · SORRY-Benchcompliance rate: 83.2 -> 69.8 (EN FT, EN eval)Improved
use more conservative LoRA scaling alpha/r=2 instead of 8 (ablation)Gemma-3-1B-IT · SORRY-Bench English eval (EN FT)compliance rate (%): 65.5 (alpha/r=8) vs 44.1 (alpha/r=2)No clear effect
fine-tune in non-English language instead of Englishsmall models · SORRY-Bench local evalcompliance increase vs EN baseline: two-thirds of cases larger than EnglishMixed
add capability (TinyMMLU) improvement as predictor of compliance changebenign multilingual fine-tuning · TinyMMLU vs SORRY-Bench ENcorrelationNo clear effect
raise language resource levelLlama-3.2-1B · SORRY-Bench local evalSpearman r: r = 0.840, p = 0.005Improved
raise language resource levelGemma-3-4B · SORRY-Bench local evalSpearman r: r = 0.225, p = 0.560No clear effect
larger safety-direction displacement at 80th percentile layerLlama models · SORRY-Bench ENcompliance changeMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy