mlsanityData to 7 Oct 2026

Are Smarter LLMs Safer? Exploring Safety-Reasoning Trade-offs in Prompting and Fine-Tuning

arXiv 2502.09673 ↗First posted 13 Feb 2025cs.CL cs.AIcited by 19

On the map: LLM reasoning → Reasoning for agents, planning and code

also in 1 more
Abstract (arXiv)

Large Language Models (LLMs) have demonstrated remarkable success across various NLP benchmarks. However, excelling in complex tasks that require nuanced reasoning and precise decision-making demands more than raw language proficiency--LLMs must reason, i.e., think logically, draw from past experiences, and synthesize information to reach conclusions and take action. To enhance reasoning abilities, approaches such as prompting and fine-tuning have been widely explored. While these methods have led to clear improvements in reasoning, their impact on LLM safety remains less understood. In this work, we investigate the interplay between reasoning and safety in LLMs. We highlight the latent safety risks that arise as reasoning capabilities improve, shedding light on previously overlooked vulnerabilities. At the same time, we explore how reasoning itself can be leveraged to enhance safety, uncovering potential mitigation strategies. By examining both the risks and opportunities in reasoning-driven LLM safety, our study provides valuable insights for developing models that are not only more capable but also more trustworthy in real-world deployments.

Key facts. Cited by 19 papers in the map (data through 7 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 45 of its 60 references that are in the map. First posted on arXiv 13 Feb 2025.

Builds on

45 of 60 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024To Err is Machine: Vulnerability Detection Challenges LLM Reasoning25 Mar

Other references · 44

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2024The Llama 3 Herd of Modelscited by 14,192
  4. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  5. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  6. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  7. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  8. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  9. 2022Large Language Models are Zero-Shot Reasonerscited by 4,323
  10. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
  11. 2022Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answeringcited by 1,881
  12. 2024HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusalcited by 1,330
32 more
  1. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,106
  2. 2024Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvementcited by 1,057
  3. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 965
  4. 2022Automatic Chain of Thought Prompting in Large Language Modelscited by 684
  5. 2023DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Modelscited by 505
  6. 2023Jailbreak and Guard Aligned Language Models with Only Few In-Context Demonstrationscited by 401
  7. 2024TrustLLM: Trustworthiness in Large Language Modelscited by 280
  8. 2024Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Modelscited by 277
  9. 2023Prompting Large Language Model for Machine Translation: A Case Studycited by 247
  10. 2022On Second Thought, Let's Not Think Step by Step! Bias and Toxicity in Zero-Shot Reasoningcited by 191
  11. 2023Bias Runs Deep: Implicit Reasoning Biases in Persona-Assigned LLMscited by 191
  12. 2023SafetyBench: Evaluating the Safety of Large Language Modelscited by 189
  13. 2022GreaseLM: Graph REASoning Enhanced Language Models for Question Answeringcited by 170
  14. 2024Marco-o1: Towards Open Reasoning Models for Open-Ended Solutionscited by 142
  15. 2023Verify-and-Edit: A Knowledge-Enhanced Chain-of-Thought Frameworkcited by 139
  16. 2023Automatic Prompt Augmentation and Selection with Chain-of-Thought from Labeled Datacited by 119
  17. 2024What is in Your Safe Data? Identifying Benign Data that Breaks Safetycited by 114
  18. 2024Making Reasoning Matter: Measuring and Improving Faithfulness of Chain-of-Thought Reasoningcited by 108
  19. 2023Question Decomposition Improves the Faithfulness of Model-Generated Reasoningcited by 107
  20. 2024Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templatescited by 94
  21. 2024OpenR: An Open Source Framework for Advanced Reasoning with Large Language Modelscited by 92
  22. 2024DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLM Jailbreakerscited by 91
  23. 2024Bias-Augmented Consistency Training Reduces Biased Reasoning in Chain-of-Thoughtcited by 37
  24. 2023Rethinking Machine Ethics -- Can LLMs Perform Moral Reasoning through the Lens of Moral Theories?cited by 36
  25. 2023Deceptive Semantic Shortcuts on Reasoning Chains: How Far Can Models Go without Hallucination?cited by 26
  26. 2024Stop Reasoning! When Multimodal LLM with Chain-of-Thought Reasoning Meets Adversarial Imagecited by 26
  27. 2023Multi-modal Latent Space Learning for Chain-of-Thought Reasoning in Language Modelscited by 24
  28. 2024Separate the Wheat from the Chaff: A Post-Hoc Approach to Safety Re-Alignment for Fine-Tuned Language Modelscited by 19
  29. 2023Retrieval-augmented Multi-modal Chain-of-Thoughts Reasoning for Large Language Modelscited by 18
  30. 2024You Know What I'm Saying: Jailbreak Attack via Implicit Referencecited by 13
  31. 2024In-Context Learning May Not Elicit Trustworthy Reasoning: A-Not-B Errors in Pretrained Language Modelscited by 8
  32. 2024Enhancing Question Answering for Enterprise Knowledge Bases using Large Language Modelscited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 9; 2026: 10.

19 papers.

Sort:

By area:

  1. 2025Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonablecited by 120
  2. 2025Leaky Thoughts: Large Reasoning Models Are Not Private Thinkerscited by 25
  3. 2025A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Modelscited by 22
  4. 2025JADES: A Universal Framework for Jailbreak Assessment via Decompositional Scoringcited by 12
  5. 2025VisCRA: A Visual Chain Reasoning Attack for Jailbreaking Multimodal Large Language Modelscited by 11
  6. 2025Refusal Falls off a Cliff: How Safety Alignment Fails in Reasoning?cited by 11
  7. 2026THINKSAFE: Self-Generated Safety Alignment for Reasoning Modelscited by 9
  8. 2025LoRA is All You Need for Safety Alignment of Reasoning LLMscited by 6
  9. 2026What Matters For Safety Alignment?cited by 6
  10. 2025When Models Outthink Their Safety: Unveiling and Mitigating Self-Jailbreak in Large Reasoning Modelscited by 5
  11. 2025Efficiency vs. Alignment: Investigating Safety and Fairness Risks in Parameter-Efficient Fine-Tuning of LLMscited by 4
  12. 2026Mitigating Safety Tax via Distribution-Grounded Refinement in Large Reasoning Modelscited by 3
7 more
  1. 2026Benchmark of Benchmarks: Unpacking Influence and Code Repository Quality in LLM Safety Benchmarkscited by 3
  2. 2026Alignment-Weighted DPO: A principled reasoning approach to improve safety alignmentcited by 2
  3. 2026Finding and Reactivating Post-Trained LLMs' Hidden Safety Mechanismscited by 2
  4. 2026Safety Training May Persist Through Helpfulness Optimization in LLM Agentscited by 1
  5. 2026SAE as a Crystal Ball: Interpretable Features Predict Cross-domain Transferability of LLMs without Trainingcited by 1
  6. 2026TrustLDM: Benchmarking Trustworthiness in Language Diffusion Modelscited by 1
  7. 2026An Empirical Measurement of Jailbreaking Evaluatorscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Safety is Not Only About Refusal: Reasoning-Enhanced Fine-tuning for Interpretable LLM Safety
  2. 2025SafeRBench: Dissecting the Reasoning Safety of Large Language Models
  3. 2025Exploring LLM Reasoning Through Controlled Prompt Variations
  4. 2025SafeLawBench: Towards Safe Alignment of Large Language Models
  5. 2024SG-Bench: Evaluating LLM Safety Generalization Across Diverse Tasks and Prompt Types
  6. 2025How Should We Enhance the Safety of Large Reasoning Models: An Empirical Study
  7. 2025Reasoning Strategies in Large Language Models: Can They Follow, Prefer, and Optimize?
  8. 2025Picky LLMs and Unreliable RMs: An Empirical Study on Safety Alignment after Instruction Tuning