Linguistic Loopholes in LLM Unlearning: From a 174-Language Benchmark to Coverage-Aware Unlearning
cs.CL, cs.AI
Submitted: 2026-09-30
Updated: 2026-09-30
Code: https://github.com/tskow99/crosslingual-unlearning-tensor
Project page: https://tskow99.github.io/crosslingual-unlearning-tensor
Terminology
Sources
- Extracting Training Data from Large Language Models
- OpenUnlearning: Accelerating LLM Unlearning via Unified Benchmarking of Methods and Metrics
- The Llama 3 Herd of Models
- Uncovering the Potential Risks in Unlearning: Danger of English-only Unlearning in Multilingual LLMs
- mu squared-Bench: A Multilingual Machine Unlearning Benchmark
- The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning
- Layer-Targeted Multilingual Knowledge Erasure in Large Language Models
- Evaluating Cross-Lingual Unlearning in Multilingual Language Models
- Analyzing Leakage of Personally Identifiable Information in Language Models
- TOFU: A Task of Fictitious Unlearning for LLMs
- Tiny Aya: Bridging Scale and Multilingual Depth
- FAME: Fictional Actors for Multilingual Erasure
- Beyond Cross-Lingual Transfer: Benchmarking Propagation Boundaries in Multilingual LLM Unlearning
- No Language Left Behind: Scaling Human-Centered Machine Translation
- Qwen3 Technical Report
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering