Aussie AI

Refusal Module

Last Updated 22 October, 2025

by David Spuler, Ph.D.

Research on Refusal Module

Research papers include:

Andy Arditi, Oscar Obeso, Aaquib111, wesg, Neel Nanda, 27th Apr 2024, Refusal in LLMs is mediated by a single direction, LessWrong, https://www.lesswrong.com/posts/jGuXSZgv6qfdhMCuJ/refusal-in-llms-is-mediated-by-a-single-direction
Maxime Labonne June 13, 2024 Uncensor any LLM with abliteration, https://huggingface.co/blog/mlabonne/abliteration
NVIDIA, June 2024, Nemotron-4 340B Technical Report, https://d1qx31qr3h6wln.cloudfront.net/publications/Nemotron_4_340B_8T_0.pdf (Architecture is decoder-only with GQA, SentencePiece tokenizer, causal attention masks, RoPE, 96 layers, 96 heads, 8 KV heads, 256,000 vocabulary, 18432 internal dimension, context window 4096, and uses squared RELU.)
Seungju Han, Kavel Rao, Allyson Ettinger, Liwei Jiang, Bill Yuchen Lin, Nathan Lambert, Yejin Choi, Nouha Dziri, 26 Jun 2024, WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs, https://arxiv.org/abs/2406.18495
Maksym Andriushchenko, Nicolas Flammarion, 16 Jul 2024, Does Refusal Training in LLMs Generalize to the Past Tense? https://arxiv.org/abs/2407.11969 Code: https://github.com/tml-epfl/llm-past-tense
Kylie Robison, Jul 20, 2024, OpenAI’s latest model will block the ‘ignore all previous instructions’ loophole, https://www.theverge.com/2024/7/19/24201414/openai-chatgpt-gpt-4o-prompt-injection-instruction-hierarchy
Xinyi Hou, Yanjie Zhao, Haoyu Wang, 3 Aug 2024, Voices from the Frontier: A Comprehensive Analysis of the OpenAI Developer Forum, https://arxiv.org/abs/2408.01687
Asir Saadat, Tasmia Binte Sogir, Md Taukir Azam Chowdhury, Syem Aziz, 16 Oct 2024, When Not to Answer: Evaluating Prompts on GPT Models for Effective Abstention in Unanswerable Math Word Problems, https://arxiv.org/abs/2410.13029
Kyle O'Brien, David Majercak, Xavier Fernandes, Richard Edgar, Jingya Chen, Harsha Nori, Dean Carignan, Eric Horvitz, Forough Poursabzi-Sangde, 18 Nov 2024, Steering Language Model Refusal with Sparse Autoencoders, https://arxiv.org/abs/2411.11296
Mohit Sewak, Dec 6, 2024, Prompt Injection Attacks on Large Language Models, https://pub.towardsai.net/prompt-injection-attacks-on-large-language-models-bd8062fa1bb7
Yue Liu, Hongcheng Gao, Shengfang Zhai, Jun Xia, Tianyi Wu, Zhiwei Xue, Yulin Chen, Kenji Kawaguchi, Jiaheng Zhang, Bryan Hooi, 30 Jan 2025, GuardReasoner: Towards Reasoning-based LLM Safeguards, https://arxiv.org/abs/2501.18492
Holistic AI Team, March 6, 2025, Anthropic’s Claude 3.7 Sonnet Jailbreaking & Red Teaming Audit: The Most Secure Model Yet? https://www.holisticai.com/blog/claude-3-7-sonnet-jailbreaking-audit
Vishnu Kabir Chhabra, Mohammad Mahdi Khalili, 5 Apr 2025, Towards Understanding and Improving Refusal in Compressed Models via Mechanistic Interpretability, https://arxiv.org/abs/2504.04215
Wojciech Zaremba, Evgenia Nitishinskaya, Boaz Barak, Stephanie Lin, Sam Toyer, Yaodong Yu, Rachel Dias, Eric Wallace, Kai Xiao, Johannes Heidecke, Amelia Glaese, 31 Jan 2025, Trading Inference-Time Compute for Adversarial Robustness, https://arxiv.org/abs/2501.18841
Hongzhe Du, Weikai Li, Min Cai, Karim Saraipour, Zimin Zhang, Himabindu Lakkaraju, Yizhou Sun, Shichang Zhang, 11 Aug 2025, How Post-Training Reshapes LLMs: A Mechanistic View on Knowledge, Truthfulness, Refusal, and Confidence, https://arxiv.org/abs/2504.02904
Haonan Zhang, Dongxia Wang, Yi Liu, Kexin Chen, Jiashui Wang, Xinlei Ying, Long Liu, Wenhai Wang, 15 Aug 2025, ORFuzz: Fuzzing the "Other Side" of LLM Safety -- Testing Over-Refusal, https://arxiv.org/abs/2508.11222
Yuan Yuan, Tina Sriskandarajah, Anna-Luisa Brakman, Alec Helyar, Alex Beutel, Andrea Vallone, Saachi Jain, 12 Aug 2025, From Hard Refusals to Safe-Completions: Toward Output-Centric Safety Training, https://arxiv.org/abs/2508.09224
Ranjie Duan, Jiexi Liu, Xiaojun Jia, Shiji Zhao, Ruoxi Cheng, Fengxiang Wang, Cheng Wei, Yong Xie, Chang Liu, Defeng Li, Yinpeng Dong, Yichi Zhang, Yuefeng Chen, Chongwen Wang, Xingjun Ma, Xingxing Wei, Yang Liu, Hang Su, Jun Zhu, Xinfeng Li, Yitong Sun, Jie Zhang, Jinzhao Hu, Sha Xu, Yitong Yang, Jialing Tao, Hui Xue, 4 Sep 2025, Oyster-I: Beyond Refusal -- Constructive Safety Alignment for Responsible Language Models, https://arxiv.org/abs/2509.01909
Md Abdullah Al Mamun, Ihsen Alouani, Nael Abu-Ghazaleh, 28 Aug 2025, Poison Once, Refuse Forever: Weaponizing Alignment for Injecting Bias in LLMs, https://arxiv.org/abs/2508.20333
Neel Jain, Aditya Shrivastava, Chenyang Zhu, Daben Liu, Alfy Samuel, Ashwinee Panda, Anoop Kumar, Micah Goldblum, Tom Goldstein, 29 Aug 2025, Refusal Tokens: A Simple Way to Calibrate Refusals in Large Language Models, https://arxiv.org/abs/2412.06748
Nirmalendu Prakash, Yeo Wei Jie, Amir Abdullah, Ranjan Satapathy, Erik Cambria, Roy Ka Wei Lee, 7 Sep 2025, Beyond I'm Sorry, I Can't: Dissecting Large Language Model Refusal, https://arxiv.org/abs/2509.09708