arXiv Computation and Language
When the Weakest Language Keeps Moving: ARMOR-DP for Multilingual BabyLM
Read the original on arXiv Computation and Language →The Flow has not summarised this story yet — read it at arXiv Computation and Language.
Related stories
arXiv Computation and Language
Oct 15
Language Exposure Matters: Distribution and Order in Multilingual BabyLM Pre-training
arXiv:2610.20102v1 Announce Type: new Abstract: We investigate two aspects of multilingual pre-training with limited data, namely how language exposure is distributed across English, Dutch, and Chine...
arXiv Machine Learning
Jul 3
HaloGuard 1.0: An Open Weights Constitutional Classifier for Multilingual AI Safety
arXiv:2607. 02079v1 Announce Type: cross Abstract: We present HaloGuard 1.
arXiv Machine Learning
Aug 17
Language-Specific Gaps in AI Safety Training Datasets
arXiv:2608. 13695v1 Announce Type: cross Abstract: Large language model providers routinely cite multilingual safety benchmarks spanning a dozen or more languages as evidence that their models are safe for non-English-speaking users.
arXiv AI
Oct 9
Obscuring Data Contamination Through Translation: Evidence from Arabic Corpora
The study investigates how exposing large language models to Arabic translations of English benchmark tasks can hide data contamination from standard English-only detection methods. By testing four instruction-tuned models with increasing Arabic exposure, the authors find that traditional probes like TS-Guessing and Min-K%++ fail to detect contamination, yet English performance still improves. They propose a new Translation-Aware Contamination Detection (TACD) method that uses cross-lingual consistency and choice reordering to reveal contamination effects that are otherwise invisible to English-only diagnostics.
arXiv AI
Jun 30
The Heterogeneous Safety Impacts of Benign Multilingual Fine-Tuning
arXiv:2606. 28843v1 Announce Type: cross Abstract: Fine-tuning a large language model is a ubiquitous method for enhancing its capability on a specific downstream task.
arXiv Computation and Language
Oct 16
PRISM: Transfer and Limits of Activation Interventions in Tool-Calling Language Models
PRISM is a framework that evaluates how activation interventions transfer across precision, languages, and model scales to mitigate unsafe tool‑calling behaviors in language models. By leveraging lower‑resource settings such as quantized NF4 labels and cross‑lingual transfer, PRISM demonstrates that alignment improvements can be achieved with minimal scenarios and that these gains transfer to held‑out languages and larger models. The study shows that auxiliary supervision from compressed models and multilingual data can effectively uncover and reduce hazardous tool‑calling behaviors across different model families.