On the Efficiency-Safety Dilemma in Large Reasoning Models
cs.CL, cs.AI
Submitted: 2026-09-20
Updated: 2026-09-20
Comments: Accepted as Main of EMNLP2026
Code: https://github.com/yangyifei729/LRM-Efficiency-Safety
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- GPT-4 Technical Report
- SliceGPT: Compress Large Language Models by Deleting Rows and Columns
- HarmLevelBench: Evaluating Harm-Level Compliance and the Impact of Quantization on Model Alignment
- PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling
- Towards Safer Large Reasoning Models by Promoting Safety Decision-Making before Chain-of-Thought Generation
- A Wolf in Sheep's Clothing: Generalized Nested Jailbreak Prompts can Fool Large Language Models Easily
- GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers
- The Llama 3 Herd of Models
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- Measuring Mathematical Problem Solving With the MATH Dataset
- Decoding Compressed Trust: Scrutinizing the Trustworthiness of Efficient LLMs Under Compression
- Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation
- Reasoning Structure Matters for Safety Alignment of Reasoning Models
- Mistral 7B
- SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities
- Investigating the Impact of Quantization Methods on the Safety and Reliability of Large Language Models
- Fine-Tuning, Quantization, and LLMs: Navigating Unintended Outcomes
- H-CoT: Hijacking the Chain-of-Thought Safety Reasoning Mechanism to Jailbreak Large Reasoning Models, Including OpenAI o1/o3, DeepSeek-R1, and Gemini 2.0 Flash Thinking
- Multi-step Jailbreaking Privacy Attacks on ChatGPT
- AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering