ICML 2025 PUT Workshop "Reasoning as an Adaptive Defense for Safety"

Talking about some conflict views and conclusions in Alignment Science.

Anthropic Techniqual Report "Reasoning Models Don't Always Say What They Think"

ACL 2025 Findings Paper "SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities"

ICLR 2025 Paper "SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal"

ICML 2025 Paper "OR-Bench: An Over-Refusal Benchmark for Large Language Models"

NIPS 2025 Paper "A StrongREJECT for Empty Jailbreaks"

NAACL 2024 Paper "XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models"