English
Related papers

Related papers: Steering Large Language Model Activations in Spars…

200 papers

Large language models (LLMs) can sometimes detect when they are being evaluated and adjust their behavior to appear more aligned, compromising the reliability of safety evaluations. In this paper, we show that adding a steering vector to an…

Computation and Language · Computer Science 2026-03-03 Tim Tian Hua , Andrew Qin , Samuel Marks , Neel Nanda

Self-destructive behaviors are linked to complex psychological states and can be challenging to diagnose. These behaviors may be even harder to identify within subcultural groups due to their unique expressions. As large language models…

Computation and Language · Computer Science 2026-05-26 Peng Wang , Xilin Tao , Siyi Yao , Jiageng Wu , Yuntao Zou , Zhuotao Tian , Libo Qin , Dagang Li

Sparse Autoencoders (SAEs) have emerged as a powerful paradigm for disentangling feature superposition in transformer-based architectures, enabling precise control via activation steering. However, the theoretical foundations of…

Machine Learning · Computer Science 2026-05-08 Yunpeng Zhou

Despite the growing reasoning capabilities of recent large language models (LLMs), their internal mechanisms during the reasoning process remain underexplored. Prior approaches often rely on human-defined concepts (e.g., overthinking,…

Computation and Language · Computer Science 2026-01-01 Zhenyu Zhang , Shujian Zhang , John Lambert , Wenxuan Zhou , Zhangyang Wang , Mingqing Chen , Andrew Hard , Rajiv Mathews , Lun Wang

Reward models (RMs) are a core component in the post-training of large language models (LLMs), serving as proxies for human preference evaluation and guiding model alignment. However, training reliable RMs under limited resources remains…

Artificial Intelligence · Computer Science 2025-11-12 Dengcan Liu , Jiahao Li , Zheren Fu , Yi Tu , Jiajun Li , Zhendong Mao , Yongdong Zhang

Sparse autoencoders (SAEs) are a promising technique for decomposing language model activations into interpretable linear features. However, current SAEs fall short of completely explaining model performance, resulting in "dark matter":…

Machine Learning · Computer Science 2025-03-26 Joshua Engels , Logan Riggs , Max Tegmark

Understanding the multilingual mechanisms of large language models (LLMs) provides insight into how they process different languages, yet this remains challenging. Existing studies often focus on individual neurons, but their polysemantic…

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

Large Vision-Language Models (LVLMs) have achieved remarkable success across cross-modal tasks but remain hindered by hallucinations, producing textual outputs inconsistent with visual content. Existing methods mitigate hallucinations but…

Computer Vision and Pattern Recognition · Computer Science 2026-04-10 Yuanhong Zhang , Zhaoyang Wang , Xin Zhang , Weizhan Zhang , Joey Tianyi Zhou

Steering vectors are a promising approach to control the behaviour of large language models. However, their underlying mechanisms remain poorly understood. While sparse autoencoders (SAEs) may offer a potential method to interpret steering…

Machine Learning · Computer Science 2024-11-14 Harry Mayne , Yushi Yang , Adam Mahdi

An unintended consequence of the vast pretraining of Large Language Models (LLMs) is the verbatim memorization of fragments of their training data, which may contain sensitive or copyrighted information. In recent years, unlearning has…

Computation and Language · Computer Science 2024-11-06 Atakan Seyitoğlu , Aleksei Kuvshinov , Leo Schwinn , Stephan Günnemann

Large language models (LLMs) have demonstrated proficiency across various natural language processing (NLP) tasks but often require additional training, such as continual pre-training and supervised fine-tuning. However, the costs…

Computation and Language · Computer Science 2024-06-07 Da Ma , Lu Chen , Pengyu Wang , Hongshen Xu , Hanqi Li , Liangtai Sun , Su Zhu , Shuai Fan , Kai Yu

Recent work on activation and latent steering has demonstrated that modifying internal representations can effectively guide large language models (LLMs) toward improved reasoning and efficiency without additional training. However, most…

Machine Learning · Computer Science 2026-01-07 Tuc Nguyen , Thai Le

Jailbreak attacks remain a persistent threat to large language model safety. We propose Context-Conditioned Delta Steering (CC-Delta), an SAE-based defense that identifies jailbreak-relevant sparse features by comparing token-level…

Cryptography and Security · Computer Science 2026-02-16 Yannick Assogba , Jacopo Cortellazzi , Javier Abad , Pau Rodriguez , Xavier Suau , Arno Blaas

Large language models (LLMs) often exhibit undesirable behaviours, such as generating untruthful or biased content. Editing their internal representations has been shown to be effective in mitigating such behaviours on top of the existing…

Computation and Language · Computer Science 2024-11-05 Yifu Qiu , Zheng Zhao , Yftah Ziser , Anna Korhonen , Edoardo M. Ponti , Shay B. Cohen

We use sparse autoencoder (SAE) feature steering to amplify Dark Triad personality traits (Machiavellianism, narcissism, and psychopathy) in Llama-3.3-70B-Instruct and evaluate the resulting behavioral changes across five psychological…

Computation and Language · Computer Science 2026-05-12 Cameron Berg , Roshni Lulla

Sparse autoencoders (SAEs) have lately been used to uncover interpretable latent features in large language models. By projecting dense embeddings into a much higher-dimensional and sparse space, learned features become disentangled and…

Machine Learning · Computer Science 2025-07-30 Viktoria Schuster

Large language models (LLMs) achieve strong reasoning performance by allocating substantial computation at inference time, often generating long and verbose reasoning traces. While recent work on efficient reasoning reduces this overhead…

Computation and Language · Computer Science 2026-04-28 Han Wang , Xiaodong Yu , Jialian Wu , Jiang Liu , Ximeng Sun , Mohit Bansal , Zicheng Liu

Understanding the internal machinations of deep Transformer-based NLP models is more crucial than ever as these models see widespread use in various domains that affect the public at large, such as industry, academia, finance, health. While…

Computation and Language · Computer Science 2026-05-13 Dan Pluth , Zachary Nicholas Houghton , Yu Zhou , Vijay K. Gurbani

Large Language Models (LLMs) exhibit significant activation sparsity, where only a subset of neurons are active for a given input. Although this sparsity presents opportunities to reduce computational cost, efficiently utilizing it requires…

Machine Learning · Computer Science 2025-07-22 Nobel Dhar , Bobin Deng , Md Romyull Islam , Xinyue Zhang , Kazi Fahim Ahmad Nasif , Kun Suo