# Safety in Large Reasoning Models: A Survey

arXiv: https://arxiv.org/abs/2504.17704 · HTML: https://mlsanity.com/paper/2504.17704/

## Key facts

- Cited by 65 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning.
- Builds on 88 of its 118 references that are in the map.
- First posted on arXiv 24 Apr 2025.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)

## Builds on (most cited first, 30 of 88)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [Let's Verify Step by Step](https://mlsanity.com/paper/2305.20050/) (2023)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code](https://mlsanity.com/paper/2403.07974/) (2024)
- [Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations](https://mlsanity.com/paper/2312.06674/) (2023)
- [A General Language Assistant as a Laboratory for Alignment](https://mlsanity.com/paper/2112.00861/) (2021)
- [Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate](https://mlsanity.com/paper/2305.19118/) (2023)
- [Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks](https://mlsanity.com/paper/2204.07705/) (2022)
- [Prompt Injection attack against LLM-integrated Applications](https://mlsanity.com/paper/2306.05499/) (2023)
- [How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs](https://mlsanity.com/paper/2401.06373/) (2024)
- [Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs](https://mlsanity.com/paper/2412.21187/) (2024)
- [WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs](https://mlsanity.com/paper/2406.18495/) (2024)
- [Great, Now Write an Article About That: The Crescendo Multi-Turn LLM Jailbreak Attack](https://mlsanity.com/paper/2404.01833/) (2024)
- [R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization](https://mlsanity.com/paper/2503.10615/) (2025)
- [DeepInception: Hypnotize Large Language Model to Be Jailbreaker](https://mlsanity.com/paper/2311.03191/) (2023)
- [Deliberative Alignment: Reasoning Enables Safer Language Models](https://mlsanity.com/paper/2412.16339/) (2024)
- [Alphazero-like Tree-Search can Guide Large Language Model Decoding and Training](https://mlsanity.com/paper/2309.17179/) (2023)
- [Recursively Summarizing Books with Human Feedback](https://mlsanity.com/paper/2109.10862/) (2021)
- [PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts](https://mlsanity.com/paper/2202.01279/) (2022)
- [Large Language Model Alignment: A Survey](https://mlsanity.com/paper/2309.15025/) (2023)
- [ShieldGemma: Generative AI Content Moderation Based on Gemma](https://mlsanity.com/paper/2407.21772/) (2024)
- [ReFT: Reasoning with Reinforced Fine-Tuning](https://mlsanity.com/paper/2401.08967/) (2024)
- [Jailbreak Attacks and Defenses Against Large Language Models: A Survey](https://mlsanity.com/paper/2407.04295/) (2024)
- [ToxicChat: Unveiling Hidden Challenges of Toxicity Detection in Real-World User-AI Conversation](https://mlsanity.com/paper/2310.17389/) (2023)
- [Hidden Killer: Invisible Textual Backdoor Attacks with Syntactic Trigger](https://mlsanity.com/paper/2105.12400/) (2021)
- [Scalable and Transferable Black-Box Jailbreaks for Language Models via Persona Modulation](https://mlsanity.com/paper/2311.03348/) (2023)
- [Chain-of-Table: Evolving Tables in the Reasoning Chain for Table Understanding](https://mlsanity.com/paper/2401.04398/) (2024)

## Built on by (most cited first, 30 of 65)

- [GuardReasoner-VL: Safeguarding VLMs via Reinforced Reasoning](https://mlsanity.com/paper/2505.11049/) (2025)
- [AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions](https://mlsanity.com/paper/2506.14697/) (2025)
- [Jigsaw-R1: A Study of Rule-based Visual Reinforcement Learning with Jigsaw Puzzles](https://mlsanity.com/paper/2505.23590/) (2025)
- [Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Models](https://mlsanity.com/paper/2508.02120/) (2025)
- [EvoTest: Evolutionary Test-Time Learning for Self-Improving Agentic Systems](https://mlsanity.com/paper/2510.13220/) (2025)
- [Leaky Thoughts: Large Reasoning Models Are Not Private Thinkers](https://mlsanity.com/paper/2506.15674/) (2025)
- [Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection](https://mlsanity.com/paper/2507.02844/) (2025)
- [SAFEPATH: Preventing Harmful Reasoning in Chain-of-Thought via Early Alignment](https://mlsanity.com/paper/2505.14667/) (2025)
- [Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets](https://mlsanity.com/paper/2506.05346/) (2025)
- [False Sense of Security: Why Probing-based Malicious Input Detection Fails to Generalize](https://mlsanity.com/paper/2509.03888/) (2025)
- [Robustness via Referencing: Defending against Prompt Injection Attacks by Referencing the Executed Instruction](https://mlsanity.com/paper/2504.20472/) (2025)
- [SafeMobile: Chain-level Jailbreak Detection and Automated Evaluation for Multimodal Mobile Agents](https://mlsanity.com/paper/2507.00841/) (2025)
- [Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills](https://mlsanity.com/paper/2506.12963/) (2025)
- [TROJail: Trajectory-Level Optimization for Multi-Turn Large Language Model Jailbreaks with Process Rewards](https://mlsanity.com/paper/2512.07761/) (2025)
- [Pushing the Limits of Safety: A Technical Report on the ATLAS Challenge 2025](https://mlsanity.com/paper/2506.12430/) (2025)
- [Large Reasoning Models Learn Better Alignment from Flawed Thinking](https://mlsanity.com/paper/2510.00938/) (2025)
- [Self-Jailbreaking: Language Models Can Reason Themselves Out of Safety Alignment After Benign Reasoning Training](https://mlsanity.com/paper/2510.20956/) (2025)
- [Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive Scoring](https://mlsanity.com/paper/2512.12069/) (2025)
- [Refusal Falls off a Cliff: How Safety Alignment Fails in Reasoning?](https://mlsanity.com/paper/2510.06036/) (2025)
- [SECA: Semantically Equivalent and Coherent Attacks for Eliciting LLM Hallucinations](https://mlsanity.com/paper/2510.04398/) (2025)
- [Safety Under Scaffolding: How Evaluation Conditions Shape Measured Safety](https://mlsanity.com/paper/2603.10044/) (2026)
- [BadThink: Triggered Overthinking Attacks on Chain-of-Thought Reasoning in Large Language Models](https://mlsanity.com/paper/2511.10714/) (2025)
- [EvoClinician: A Self-Evolving Agent for Multi-Turn Medical Diagnosis via Test-Time Evolutionary Learning](https://mlsanity.com/paper/2601.22964/) (2026)
- [Beyond Safe Answers: A Benchmark for Evaluating True Risk Awareness in Large Reasoning Models](https://mlsanity.com/paper/2505.19690/) (2025)
- [Efficient Reasoning via Chain of Unconscious Thought](https://mlsanity.com/paper/2505.19756/) (2025)
- [UnsafeChain: Enhancing Reasoning Model Safety via Hard Cases](https://mlsanity.com/paper/2507.21652/) (2025)
- [What Matters For Safety Alignment?](https://mlsanity.com/paper/2601.03868/) (2026)
- [EMO-R3: Reflective Reinforcement Learning for Emotional Reasoning in Multimodal Large Language Models](https://mlsanity.com/paper/2602.23802/) (2026)
- [Thought Purity: A Defense Framework For Chain-of-Thought Attack](https://mlsanity.com/paper/2507.12314/) (2025)
- [ReasoningGuard: Safeguarding Large Reasoning Models with Inference-time Safety Aha Moments](https://mlsanity.com/paper/2508.04204/) (2025)

## Abstract (arXiv)

Large Reasoning Models (LRMs) have exhibited extraordinary prowess in tasks like mathematics and coding, leveraging their advanced reasoning capabilities. Nevertheless, as these capabilities progress, significant concerns regarding their vulnerabilities and safety have arisen, which can pose challenges to their deployment and application in real-world settings. This paper presents a comprehensive survey of LRMs, meticulously exploring and summarizing the newly emerged safety risks, attacks, and defense strategies. By organizing these elements into a detailed taxonomy, this work aims to offer a clear and structured understanding of the current safety landscape of LRMs, facilitating future research and development to enhance the security and reliability of these powerful models.