EU AI Act Safety Evaluations

community

AI & ML interests

Redteaming; Safety

authored 2 papers 5 months ago

Multilingual Refusal Alignment for Safer Large Language Models

Paper • 2606.07535 • Published Apr 24

Reassessing High-Performing LLMs on Polish Medical Exams: True Competence or Bias-Driven Performance?

Paper • 2606.12250 • Published Jun 10

authored 2 papers 8 months ago

Evaluating Self-Supervised Learning in Medical Imaging: A Benchmark for Robustness, Generalizability, and Multi-Domain Impact

Paper • 2412.19124 • Published Dec 26, 2024

Half-Truths Break Similarity-Based Retrieval

Paper • 2602.23906 • Published Feb 27 • 6

authored 5 papers over 1 year ago

PL-Guard: Benchmarking Language Model Safety for Polish

Paper • 2506.16322 • Published Jun 19, 2025 • 1

ConECT Dataset: Overcoming Data Scarcity in Context-Aware E-Commerce MT

Paper • 2506.04929 • Published Jun 5, 2025 • 2

Aurora-M: The First Open Source Multilingual Language Model Red-teamed according to the U.S. Executive Order

Paper • 2404.00399 • Published Mar 30, 2024 • 42

BigBIO: A Framework for Data-Centric Biomedical Natural Language Processing

Paper • 2206.15076 • Published Jun 30, 2022 • 5

CSMeD: Bridging the Dataset Gap in Automated Citation Screening for Systematic Literature Reviews

Paper • 2311.12474 • Published Nov 21, 2023 • 1

authored a paper over 1 year ago

ARLBench: Flexible and Efficient Benchmarking for Hyperparameter Optimization in Reinforcement Learning

Paper • 2409.18827 • Published Sep 27, 2024