English
Related papers

Related papers: Sparse Feature Coactivation Reveals Causal Semanti…

200 papers

Sparse autoencoders (SAEs) have emerged as a promising approach in language model interpretability, offering unsupervised extraction of sparse features. For interpretability methods to succeed, they must identify abstract features across…

Machine Learning · Computer Science 2025-09-08 Lovis Heindrich , Philip Torr , Fazl Barez , Veronika Thost

Large language models (LLMs) acquire most of their factual knowledge during the pre-training stage, through next token prediction. Subsequent stages of post-training often introduce new facts outwith the parametric knowledge, giving rise to…

Computation and Language · Computer Science 2026-04-30 Dimitris Dimakopoulos , Shay B. Cohen , Ioannis Konstas

As large language models (LLMs) are increasingly deployed in real-world applications, the need to selectively remove unwanted knowledge while preserving model utility has become paramount. Recent work has explored sparse autoencoders (SAEs)…

Computation and Language · Computer Science 2026-04-28 Tomer Ashuach , Dana Arad , Aaron Mueller , Martin Tutek , Yonatan Belinkov

Sparse Autoencoders (SAEs) extract features from LLM internal activations, meant to correspond to interpretable concepts. A core SAE training hyperparameter is L0: how many SAE features should fire per token on average. Existing work…

Machine Learning · Computer Science 2025-12-08 David Chanin , Adrià Garriga-Alonso

Large language models (LLMs) are now ubiquitous in user-facing applications, yet they still generate undesirable toxic outputs, including profanity, vulgarity, and derogatory remarks. Although numerous detoxification methods exist, most…

Computation and Language · Computer Science 2025-10-24 Agam Goyal , Vedant Rathi , William Yeh , Yian Wang , Yuen Chen , Hari Sundaram

Sparse Autoencoders (SAEs) have demonstrated significant promise in interpreting the hidden states of language models by decomposing them into interpretable latent directions. However, training and interpreting SAEs at scale remains…

Machine Learning · Computer Science 2025-12-23 Vadim Kurochkin , Yaroslav Aksenov , Daniil Laptev , Daniil Gavrilov , Nikita Balagansky

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

Sparse autoencoders (SAEs) are now standard tools for decomposing language model activations into interpretable features, and automated interpretability pipelines routinely assign each feature a short natural-language explanation. Existing…

Machine Learning · Computer Science 2026-05-14 Jordan F. McCann

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

Sparse activation, which selectively activates only an input-dependent set of neurons in inference, is a useful technique to reduce the computing cost of Large Language Models (LLMs) without retraining or adaptation efforts. However,…

Computation and Language · Computer Science 2024-06-12 Jifeng Song , Kai Huang , Xiangyu Yin , Boyuan Yang , Wei Gao

Sparse autoencoders (SAEs) are a popular method for decomposing Large Langage Models (LLM) activations into interpretable latents. However, due to their substantial training cost, most academic research uses open-source SAEs which are only…

Machine Learning · Computer Science 2025-06-13 Patrick Leask , Neel Nanda , Noura Al Moubayed

Large language models (LLMs) can store a significant amount of factual knowledge in their parameters. However, their parametric knowledge may conflict with the information provided in the context -- this phenomenon, known as…

Computation and Language · Computer Science 2025-02-11 Yu Zhao , Alessio Devoto , Giwon Hong , Xiaotang Du , Aryo Pradipta Gema , Hongru Wang , Xuanli He , Kam-Fai Wong , Pasquale Minervini

We investigate feature universality in Gemma-2 language models (Gemma-2-2B and Gemma-2-9B), asking whether models with a four-fold difference in scale still converge on comparable internal concepts. Using the Sparse Autoencoder (SAE)…

Computation and Language · Computer Science 2025-08-01 Daniel Son , Sanjana Rathore , Andrew Rufail , Adrian Simon , Daniel Zhang , Soham Dave , Cole Blondin , Kevin Zhu , Sean O'Brien

While large language models (LLMs) have demonstrated strong capability in structured prediction tasks such as semantic parsing, few amounts of research have explored the underlying mechanisms of their success. Our work studies different…

Computation and Language · Computer Science 2023-02-01 Daking Rai , Yilun Zhou , Bailin Wang , Ziyu Yao

Large language models (LLMs) exhibit a wide range of capabilities, including mathematical reasoning, code generation, and linguistic behaviors. We show that many capabilities are highly localized to small subsets of attention heads within…

Computation and Language · Computer Science 2026-03-05 Anna Bair , Yixuan Even Xu , Mingjie Sun , J. Zico Kolter

Sparse autoencoders (SAEs) are a promising approach for uncovering interpretable features in large language models (LLMs). While several automated evaluation methods exist for SAEs, most rely on external LLMs. In this work, we introduce…

Computation and Language · Computer Science 2025-09-30 Alex Gulko , Yusen Peng , Sachin Kumar

Large Language Models (LLMs) excel at capturing latent semantics and contextual relationships across diverse modalities. However, in modeling user behavior from sequential interaction data, performance often suffers when such semantic…

Computation and Language · Computer Science 2025-10-22 Mahsa Valizadeh , Xiangjue Dong , Rui Tuo , James Caverlee

Sparse autoencoders (SAEs) offer a natural path toward comparable explanations across different representation spaces. However, current SAEs are trained per modality, producing dictionaries whose features are not directly understandable and…

Computer Vision and Pattern Recognition · Computer Science 2026-02-10 Difei Gu , Yunhe Gao , Gerasimos Chatzoudis , Zihan Dong , Guoning Zhang , Bangwei Guo , Yang Zhou , Mu Zhou , Dimitris Metaxas

Sparse autoencoders (SAEs) have been used widely to decompose and interpret neural network activations, especially those of transformer language models. One key issue with SAEs is their inability to directly model multidimensional features.…

Machine Learning · Computer Science 2026-05-12 Collin Francel

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