English
Related papers

Related papers: Refusal in LLMs is an Affine Function

200 papers

To circumvent the alignment of large language models (LLMs), current optimization-based adversarial attacks usually craft adversarial prompts by maximizing the likelihood of a so-called affirmative response. An affirmative response is a…

Fine-grained steering of language model outputs is essential for safety and reliability. Prompting and finetuning are widely used to achieve these goals, but interpretability researchers have proposed a variety of representation-based…

Computation and Language · Computer Science 2025-03-05 Zhengxuan Wu , Aryaman Arora , Atticus Geiger , Zheng Wang , Jing Huang , Dan Jurafsky , Christopher D. Manning , Christopher Potts

In this paper, we investigate the degree to which fine-tuning in Large Language Models (LLMs) effectively mitigates versus merely conceals undesirable behavior. Through the lens of semi-realistic role-playing exercises designed to elicit…

Computation and Language · Computer Science 2024-07-01 Florin Pop , Judd Rosenblatt , Diogo Schwerz de Lucena , Michael Vaiana

Masked diffusion language models (MDLMs) generate text via iterative masked-token denoising, enabling mask-parallel decoding and distinct controllability and efficiency tradeoffs from autoregressive LLMs. Yet, efficient representation-level…

Computation and Language · Computer Science 2026-03-31 Adi Shnaidman , Erin Feiglin , Osher Yaari , Efrat Mentel , Amit Levi , Raz Lapid

Vision-language models (VLMs) are capable of recognizing unseen actions. However, existing VLMs lack intrinsic understanding of procedural action concepts. Hence, they overfit to fixed labels and are not invariant to unseen action synonyms.…

Computer Vision and Pattern Recognition · Computer Science 2024-11-26 Reza Ghoddoosian , Nakul Agarwal , Isht Dwivedi , Behzad Darisuh

Refusal on harmful prompts is a key safety behaviour in instruction-tuned large language models (LLMs), yet the internal causes of this behaviour remain poorly understood. We study two public instruction-tuned models, Gemma-2-2B-IT and…

Computation and Language · Computer Science 2026-04-29 Nirmalendu Prakash , Yeo Wei Jie , Amir Abdullah , Ranjan Satapathy , Erik Cambria , Roy Ka Wei Lee

Controlling the behavior of large language models (LLMs) at inference time is essential for aligning outputs with human abilities and safety requirements. \emph{Activation steering} provides a lightweight alternative to prompt engineering…

Artificial Intelligence · Computer Science 2026-01-30 Diaoulé Diallo , Katharina Dworatzyk , Sophie Jentzsch , Peer Schütt , Sabine Theis , Tobias Hecking

We introduce Mechanistic Error Reduction with Abstention (MERA), a principled framework for steering language models (LMs) to mitigate errors through selective, adaptive interventions. Unlike existing methods that rely on fixed, manually…

Machine Learning · Computer Science 2025-10-16 Anna Hedström , Salim I. Amoukou , Tom Bewley , Saumitra Mishra , Manuela Veloso

Large reasoning models (LRMs) generate chain-of-thought (CoT) traces before producing final outputs, introducing a dynamic internal state that may complicate control mechanisms such as refusal. Unlike instruction-tuned LLMs, where refusal…

Artificial Intelligence · Computer Science 2026-05-27 Kia-Jüng Yang , Dominik Meier , Jiachen Zhao , Terry Ruas , Bela Gipp

Safety alignment is crucial for large language models (LLMs) to resist malicious instructions but often results in over-refusals, where benign prompts are unnecessarily rejected, impairing user experience and model utility. We introduce…

Machine Learning · Computer Science 2025-07-08 Mahavir Dabas , Si Chen , Charles Fleming , Ming Jin , Ruoxi Jia

Language models (LMs) can be directed to perform target tasks by using labeled examples or natural language prompts. But selecting examples or writing prompts for can be challenging--especially in tasks that involve unusual edge cases,…

Computation and Language · Computer Science 2023-10-19 Belinda Z. Li , Alex Tamkin , Noah Goodman , Jacob Andreas

Large Language Model (LLM) deployment requires guiding the LLM to recognize and not answer unsafe prompts while complying with safe prompts. Previous methods for achieving this require adjusting model weights along with other expensive…

Machine Learning · Computer Science 2025-11-04 Samaksh Bhargav , Zining Zhu

Multimodal large language models (MLLMs) may memorize sensitive cross-modal information during pretraining, making machine unlearning (MU) crucial. Existing methods typically evaluate unlearning effectiveness based on output deviations,…

Computation and Language · Computer Science 2026-05-18 Jiahui Guang , Yingjie Zhu , Cuiyun Gao , Haiyan Wang , Jing Li , Di Shao , Zhaoquan Gu

A central piece in enabling intelligent agentic behavior in foundation models is to make them capable of introspecting upon their behavior, reasoning, and correcting their mistakes as more computation or interaction is available. Even the…

Machine Learning · Computer Science 2024-07-29 Yuxiao Qu , Tianjun Zhang , Naman Garg , Aviral Kumar

Large Language Models (LLMs) often generate inconsistent responses when prompted with semantically equivalent paraphrased inputs. Recently, activation steering, a technique that modulates LLMs' behaviours by adjusting their latent…

Computation and Language · Computer Science 2025-01-23 Jingyuan Yang , Rongjun Li , Weixuan Wang , Ziyu Zhou , Zhiyong Feng , Wei Peng

Conventional mechanical design follows an iterative process in which initial concepts are refined through cycles of expert assessment and resource-intensive Finite Element Method (FEM) analysis to meet performance goals. While machine…

Machine Learning · Computer Science 2025-05-02 Yayati Jadhav , Amir Barati Farimani

Recent large language models (LLM) are leveraging human feedback to improve their generation quality. However, human feedback is costly to obtain, especially during inference. In this work, we propose LLMRefine, an inference time…

Computation and Language · Computer Science 2024-10-28 Wenda Xu , Daniel Deutsch , Mara Finkelstein , Juraj Juraska , Biao Zhang , Zhongtao Liu , William Yang Wang , Lei Li , Markus Freitag

Effective and reliable control over large language model (LLM) behavior is a significant challenge. While activation steering methods, which add steering vectors to a model's hidden states, are a promising approach, existing techniques…

Machine Learning · Computer Science 2025-04-03 Samuel Soo , Chen Guang , Wesley Teng , Chandrasekaran Balaganesh , Tan Guoxian , Yan Ming

Large language models (LLMs) are known to produce varying responses depending on prompt phrasing, indicating that subtle guidance in phrasing can steer their answers. However, the impact of this framing bias on LLM-based evaluation, where…

Computation and Language · Computer Science 2026-01-21 Yerin Hwang , Dongryeol Lee , Taegwan Kang , Minwoo Lee , Kyomin Jung

Steering, or direct manipulation of internal activations to guide LLM responses toward specific semantic concepts, is emerging as a promising avenue for both understanding how semantic concepts are stored within LLMs and advancing LLM…

Machine Learning · Computer Science 2026-02-03 Parmida Davarmanesh , Ashia Wilson , Adityanarayanan Radhakrishnan