No article text was recorded for this version.
multilingual safety gap
Version 1 of 136 papers49 conceptsfrozen 2026-10-04
Papers in this version
- Blog: The State of Multilingual LLM Safety Research: From Measuring The Language Gap To Mitigating It
- Blog: PolyGuard: A Multilingual Safety Moderation Tool for 17 Languages
- Blog: The Heterogeneous Safety Impacts of Benign Multilingual Fine-Tuning
- Blog: FrenchToxicityPrompts: a Large Benchmark for Evaluating and Mitigating Toxicity in French Texts
- Blog: SafetyPrompts: A Systematic Review of Open Datasets for Evaluating and Improving Large Language Model Safety
- Blog: Sandwich attack: Multi-language Mixture Adaptive Attack on LLMs
- Blog: PolygloToxicityPrompts: Multilingual Evaluation of Neural Toxic Degeneration in Large Language Models
- Blog: Enhancing LLM Safety through a Theoretical Minimax Game Lens
- Blog: Safety Devolution in AI Agents
- Blog: From One to Many: Expanding the Scope of Toxicity Mitigation in Language Models
- Blog: SafetyBench: Evaluating the Safety of Large Language Models
- Blog: SocialHarmBench: Revealing LLM Vulnerabilities to Socially Harmful Requests
- Blog: Measuring and Analyzing Search Engine Poisoning of Linguistic Collisions
- Blog: LLMS LOST IN TRANSLATION: M-ALERT UNCOVERS CROSS-LINGUISTIC SAFETY GAPS
- Blog: Decoding Hate: Exploring Language Models' Reactions to Hate Speech
- Blog: AEGIS2.0: A Diverse AI Safety Dataset and Risks Taxonomy for Alignment of LLM Guardrails
- Blog: A Chinese Dataset for Evaluating the Safeguards in Large Language Models
- Blog: STAR: SocioTechnical Approach to Red Teaming Language Models
- Blog: Tricking LLMs into Disobedience: Formalizing, Analyzing, and Detecting Jailbreaks
- Blog: CRiskEval: A Chinese Multi-Level Risk Evaluation Benchmark Dataset for Large Language Models
- Blog: Preference Tuning For Toxicity Mitigation Generalizes Across Languages
- Blog: Arabic Dataset for LLM Safeguard Evaluation
- Blog: MultiParaDetox: Extending Text Detoxification with Parallel Data to New Languages
- Blog: Toxicity in Multilingual Machine Translation at Scale
- Blog: AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents
- Blog: GuardBench: A Large-Scale Benchmark for Guardrail Models
- Blog: LLM Scheming Inversely Scales with Pretraining Language Coverage
- Blog: HISTOIRESMORALES: A French Dataset for Assessing Moral Alignment
- Blog: Preserving Safety in Fine-Tuned Large Language Models: A Systematic Evaluation and Mitigation Strategy
- Blog: Compromesso! Italian Many-Shot Jailbreaks undermine the safety of Large Language Models
- Blog: Dynamics of an idiostyle of a Russian suicidal blogger
- Blog: MLLMGuard: A Multi-dimensional Safety Evaluation Suite for Multimodal Large Language Models
- Blog: WildGuard: Open One-stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs
- Blog: ASSERT: Automated Safety Scenario Red Teaming for Evaluating the Robustness of Large Language Models
- Blog: Proceedings of the First Workshop on Multilingual Counterspeech Generation
- Blog: The Missing Red Line: How Commercial Pressure Erodes AI Safety Boundaries
History & citation
Current · Permanent link
- This direction split from Cross-Lingual Safety Gaps in Large Language Models (97% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_multilingual_safety_gap_361b417f,
title = {multilingual safety gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/multilingual-safety-gap/rdr_361b417f8f5d4744883c1f1086e2252b},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}This exact version (version 1, frozen text):
@misc{lacuna_multilingual_safety_gap_361b417f_v1,
title = {multilingual safety gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/multilingual-safety-gap/rdr_361b417f8f5d4744883c1f1086e2252b/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}