ACTR: Aligning Thoughts and Responses for Multilingual Safety in Reasoning LLMs
cs.AI
Submitted: 2026-09-29
Updated: 2026-09-30
Sources
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- ReasoningShield: Safety Detection over Reasoning Traces of Large Reasoning Models
- Constitutional AI: Harmlessness from AI Feedback
- Deliberative Alignment: Reasoning Enables Safer Language Models
- Safety Reasoning with Guidelines
- Low-Resource Languages Jailbreak GPT-4
- Representation Engineering: A Top-Down Approach to AI Transparency
- Qwen3 Technical Report
- Gemma 4 Technical Report
- OpenAI o1 System Card
- Universal and Transferable Adversarial Attacks on Aligned Language Models
- Language Models are Multilingual Chain-of-Thought Reasoners
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- MPO: Multilingual Safety Alignment via Reward Gap Optimization
- GUARD-SLM: Token Activation-Based Defense Against Jailbreak Attacks for Small Language Models
- Measuring Massive Multitask Language Understanding
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems