English
Related papers

Related papers: Activation Steering for Aligned Open-ended Generat…

200 papers

This paper presents a comprehensive empirical study on the safety alignment capabilities. We evaluate what matters for safety alignment in LLMs and LRMs to provide essential insights for developing more secure and reliable AI systems. We…

Computation and Language · Computer Science 2026-02-25 Xing Li , Hui-Ling Zhen , Lihao Yin , Xianzhi Yu , Zhenhua Dong , Mingxuan Yuan

A popular approach to post-training control of large language models (LLMs) is the steering of intermediate latent representations. Namely, identify a well-chosen direction depending on the task at hand and perturbs representations along…

Machine Learning · Computer Science 2026-02-04 Magamed Taimeskhanov , Samuel Vaiter , Damien Garreau

The field of large language models (LLMs) has grown rapidly in recent years, driven by the desire for better efficiency, interpretability, and safe use. Building on the novel approach of "activation engineering," this study explores…

Computation and Language · Computer Science 2025-08-26 Rumi Allbert , James K. Wiles , Vlad Grankovsky

When language model agents tackle complex software engineering tasks, they often degrade over long trajectories, which we define as *agent drift*. We focus on two recurring failure modes *overthinking* and *overacting*, i.e., where the…

Artificial Intelligence · Computer Science 2026-05-08 Yuan Sui , Yulin Chen , Yibo Li , Xue Jiang , Yufei He , Yihong Dong , Xiaoxin He , Tianyu Gao , Bryan Hooi

Large Language Models (LLMs) achieve remarkable performance through pretraining on extensive data. This enables efficient adaptation to diverse downstream tasks. However, the lack of interpretability in their underlying mechanisms limits…

Computation and Language · Computer Science 2025-06-03 Xintong Wang , Jingheng Pan , Liang Ding , Longyue Wang , Longqin Jiang , Xingshan Li , Chris Biemann

As large language models (LLMs) become widely deployed, concerns about their safety and alignment grow. An approach to steer LLM behavior, such as mitigating biases or defending against jailbreaks, is to identify which parts of a prompt…

Computation and Language · Computer Science 2025-05-20 Kenza Amara , Rita Sevastjanova , Mennatallah El-Assady

Recent advancements in large language models (LLMs) have accelerated progress toward artificial general intelligence, yet their potential to generate harmful content poses critical safety challenges. Existing alignment methods often…

Computation and Language · Computer Science 2025-10-08 Kehua Feng , Keyan Ding , Yuhao Wang , Menghan Li , Fanjunduo Wei , Xinda Wang , Qiang Zhang , Huajun Chen

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 vision-language models (LVLMs) have achieved remarkable progress in vision-language reasoning tasks, yet ensuring their safety remains a critical challenge. Recent input-side defenses detect unsafe images with CLIP and prepend safety…

Computer Vision and Pattern Recognition · Computer Science 2026-03-02 Xingyu Zhu , Beier Zhu , Junfeng Fang , Shuo Wang , Yin Zhang , Xiang Wang , Xiangnan He

Recent work has shown that LLMs can sometimes detect when steering vectors are injected into their residual stream and identify the injected concept -- a phenomenon termed "introspective awareness." We investigate the mechanisms underlying…

Machine Learning · Computer Science 2026-05-18 Uzay Macar , Li Yang , Atticus Wang , Peter Wallich , Emmanuel Ameisen , Jack Lindsey

With the rapid advancement of large language models (LLMs), their deployment in real-world applications has become increasingly widespread. LLMs are expected to deliver robust performance across diverse tasks, user preferences, and…

Computation and Language · Computer Science 2025-11-21 Wei Xia , Zhi-Hong Deng

Representation Misdirection for Unlearning (RMU), which steers model representation in the intermediate layer to a target random representation, is an effective method for large language model (LLM) unlearning. Despite its high performance,…

Computation and Language · Computer Science 2025-02-07 Dang Huu-Tien , Trung-Tin Pham , Hoang Thanh-Tung , Naoya Inoue

The emergence of Large Reasoning Models (LRMs) introduces a new paradigm of explicit reasoning, enabling remarkable advances yet posing unique risks such as reasoning manipulation and information leakage. To mitigate these risks, current…

Artificial Intelligence · Computer Science 2026-02-03 Jingnan Zheng , Jingjun Xu , Yanzhen Luo , Chenhang Cui , Gelei Deng , Zhenkai Liang , Xiang Wang , An Zhang , Tat-Seng Chua

While generative robot policies have demonstrated significant potential in learning complex, multimodal behaviors from demonstrations, they still exhibit diverse failures at deployment-time. Policy steering offers an elegant solution to…

Robotics · Computer Science 2025-05-05 Yilin Wu , Ran Tian , Gokul Swamy , Andrea Bajcsy

Steering vectors offer a training-free mechanism for controlling reasoning behaviors in large language models, but constructing effective vectors requires identifying genuine behavioral signals in the model's hidden states. For behaviors…

Computation and Language · Computer Science 2026-04-03 Haomin Zhuang , Hojun Yoo , Xiaonan Luo , Kehan Guo , Xiangliang Zhang

Speculative decoding accelerates language model inference by separating generation into fast drafting and parallel verification. Its main limitation is drafter-verifier misalignment, which limits token acceptance and reduces overall…

Machine Learning · Computer Science 2025-11-14 Frédéric Berdoz , Peer Rheinboldt , Roger Wattenhofer

Recent advancements in language models (LMs) have marked a shift toward the growing importance of post-training. Yet, post-training approaches such as supervised fine-tuning (SFT) do not guarantee the effective use of knowledge acquired…

Computation and Language · Computer Science 2025-10-30 Chunyuan Deng , Ruidi Chang , Hanjie Chen

Masked diffusion language models (MDLMs) have emerged as a promising alternative to dominant autoregressive approaches. Although they achieve competitive performance on several tasks, a substantial gap remains in open-ended text generation.…

Computation and Language · Computer Science 2026-02-02 Mengyu Ye , Ryosuke Takahashi , Keito Kudo , Jun Suzuki

Activation steering presupposes that task-relevant behaviors correspond to linear directions in activation space -- directions that should both steer the model and be readable along the unembedding. Function vectors (FVs), extracted as mean…

Machine Learning · Computer Science 2026-05-12 Mohammed Suhail B Nadaf

Safety alignment in large language models (LLMs) is achieved through fine-tuning mechanisms that regulate neuron activations to suppress harmful content. In this work, we propose a novel approach to induce disalignment by identifying and…

Machine Learning · Computer Science 2025-05-01 Yi Zhou , Wenpeng Xing , Dezhang Kong , Changting Lin , Meng Han