Elamur Research⌕
☰
home / papers / 2026 / From Weights to Activations: Is Steering the Next Frontier of Adaptation?

From Weights to Activations: Is Steering the Next Frontier of Adaptation?

Paper · 2026 · Simon Ostermann, Daniil Gurgurov, Tanja Baeumel et al.

Summary

Steering means changing a language model's behavior by editing its internal activations during inference. It has mostly been treated as an interpretability tool, so it is rarely compared with established ways of adapting LLMs. Ostermann et al. argue in this position paper that steering is itself an adaptation paradigm, on par with weight-based methods (full fine-tuning, RLHF, PEFT) and input-based methods (prompting, ICL).

To support this, they define eight functional criteria drawn from prior surveys. They then rate each family of methods against these criteria in a qualitative way, using results already reported in the literature. They run no new experiments.

The analysis places steering at a distinct point in the design space. It acts on activations rather than weights or inputs, so its changes are targeted, cheap to apply and easy to undo. The authors use this to propose a unified taxonomy of adaptation. They also give a roadmap for treating steering as a standard option rather than an experimental add-on.

Key points

  • Eight criteria: Reliability, Generalization, Specificity, Compute Efficiency, Data Efficiency, Composability, Usability and Reversibility. Each method gets a +, 0 or – rating in Table 1.
  • Methods compared:
    • FT: fine-tuning and RLHF
    • PEFT: adapters, soft prompt tuning and LoRA
    • Prompting: prompting and ICL
    • Steering, in three families: difference-based, optimization-based and dictionary-based
  • Conceptual picture (Figure 1):
    • Fine-tuning reshapes the behavior landscape defined by the weights.
    • Prompting changes the starting point of the model's trajectory.
    • Steering deflects that trajectory by adding vectors to activations.
  • Roadmap:
    • standardized benchmarks against FFT and PEFT
    • tooling comparable to Hugging Face PEFT, such as vector libraries and recipes for choosing layers
    • treating steering as a design choice from the start of a project
    • coverage of steering in NLP curricula
  • Stated limitations:
    • Coverage of steering methods is selective.
    • The ratings in Table 1 are indicative only.
    • Steering relies on an assumption of approximately linear representations, which may not hold everywhere.
    • Steering raises risks around accountability and entangled side effects.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Saarland UniversitySaarbrücken · Germany148 works · 18,287 citations🎓Ludwig-Maximilians-Universität MünchenMunich · Germany107 works · 56,478 citations🔬German Research Centre for Artificial IntelligenceKaiserslautern · Germany77 works · 29,479 citations🎓Munich Center for Machine LearningMunich · Germany57 works · 1,094 citations🎓Technological University DublinDublin · Ireland5 works · 188 citations

Also on the byline: Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

Cite

Cite as Simon Ostermann, Daniil Gurgurov, Tanja Baeumel, Michael A. Hedderich, Sebastian Lapuschkin, Wojciech Samek and Vera Schmitt. 2026. From Weights to Activations: Is Steering the Next Frontier of Adaptation?. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.1377

@inproceedings{ostermann2026weights,
  title = {{From Weights to Activations: Is Steering the Next Frontier of Adaptation?}},
  author = {Simon Ostermann, Daniil Gurgurov, Tanja Baeumel, Michael A. Hedderich, Sebastian Lapuschkin, Wojciech Samek and Vera Schmitt},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1377},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1377}
}

References

132 references, 50 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Guillaume Alain and Yoshua Bengio. 2016. Understanding intermediate layers using linear classifier probes. arXiv preprint arXiv:1610.01644. find
  2. DM Anisuzzaman, Jeffrey G Malins, Paul A Friedman, and Zachi I Attia. 2025. Fine-tuning large language models for specialized use cases. Mayo Clinic Proceedings: Digital Health, 3(1):100184. find
  3. Dana Arad, Aaron Mueller, and Yonatan Belinkov. 2025. SAEs are good for steering – if you select the right features. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 10252–10270, Suzhou, China. Association for Computational Linguistics.
  4. Andy Arditi, Oscar Obeso, Aaquib Syed, Daniel Paleka, Nina Panickssery, Wes Gurnee, and Neel Nanda. 2024. Refusal in language models is mediated by a single direction. Advances in Neural Information Processing Systems, 37:136037–136083.
  5. Akari Asai, Mohammadreza Salehi, Matthew Peters, and Hannaneh Hajishirzi. 2022. ATTEMPT: Parameter-efficient multi-task tuning via attentional mixtures of soft prompts. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 6655–6672, Abu Dhabi, United Arab…
  6. Trapit Bansal, Salaheddin Alzubi, Tong Wang, Jay-Yoon Lee, and Andrew McCallum. 2022. Meta-adapters: Parameter efficient few-shot fine-tuning through meta-learning. In Proceedings of the First International Conference on Automated Machine Learning, volume 188 of Proceedings of Machine Learning Resea… find
  7. David Bau, Bolei Zhou, Aditya Khosla, Aude Oliva, and Antonio Torralba. 2017. Network dissection: Quantifying interpretability of deep visual representations. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). find
  8. Farima Fatahi Bayat, Xin Liu, H Jagadish, and Lu Wang. 2024. Enhanced language model truthfulness with learnable intervention and uncertainty expression. In Findings of the Association for Computational Linguistics: ACL 2024, pages 12388–12400. find
  9. Reza Bayat, Ali Rahimi-Kalahroudi, Mohammad Pezeshki, Sarath Chandar, and Pascal Vincent. 2025. Steering large language model activations in sparse spaces. In Second Conference on Language Modeling. find
  10. Robert Belanec, Simon Ostermann, Ivan Srba, and Maria Bielikova. 2025. Task prompt vectors: Effective initialization through multi-task soft prompt transfer. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pages 77–94. Springer. find
  11. Shai Ben-David, John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira, and Jennifer Wortman Vaughan. 2010. A theory of learning from different domains. Machine learning, 79(1):151–175. find
  12. Tolga Bolukbasi, Kai-Wei Chang, James Y Zou, Venkatesh Saligrama, and Adam T Kalai. 2016. Man is to computer programmer as woman is to homemaker? debiasing word embeddings. Advances in neural information processing systems, 29.
  13. Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, Erik Brynjolfsson, Shyamal Buch, Dallas Card, Rodrigo Castellon, Niladri Chatterji, Annie Chen, Kathleen Creel, Jared Quincy Davis, Dora De… — On the Opportunities and Risks of Foundation Models
  14. Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M. Ziegler, Jeffrey Wu, Clemens Winter,… — Language Models are Few-Shot Learners
  15. Anwoy Chatterjee, H S V N S Kowndinya Renduchintala, Sumit Bhatia, and Tanmoy Chakraborty. 2024. POSIX: A prompt sensitivity index for large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 14550– 14565, Miami, Florida, USA. Association for Computation… find
  16. Arnav Chavan, Zhuang Liu, Deepak Gupta, Eric Xing, and Zhiqiang Shen. 2023. One-for-all: Generalized lora for parameter-efficient fine-tuning. arXiv preprint arXiv:2306.07967. — [arXiv:2306.07967](https://arxiv.org/abs/2306.07967) find
  17. Xiaojun Chen, Ting Liu, Philippe Fournier-Viger, Bowen Zhang, Guodong Long, and Qin Zhang. 2024. A fine-grained self-adapting prompt learning approach for few-shot learning with pre-trained language models. Knowledge-Based Systems, 299:111968. find
  18. Cheng-Ting Chou, George Liu, Jessica Sun, Cole Blondin, Kevin Zhu, Vasu Sharma, and Sean O’Brien. 2025. Causal language control in multilingual transformers via sparse feature steering. Preprint, arXiv:2507.13410. find
  19. Paul F Christiano, Jan Leike, Tom Brown, Miljan Martic, Shane Legg, and Dario Amodei. 2017. Deep reinforcement learning from human preferences. Advances in neural information processing systems, 30. find
  20. Daniel Commey. 2026. When" better" prompts hurt: Evaluation-driven iteration for llm applications. arXiv preprint arXiv:2601.22025. — [arXiv:2601.22025](https://arxiv.org/abs/2601.22025) find
  21. Hoagy Cunningham, Aidan Ewart, Logan Riggs Smith, Robert Huben, and Lee Sharkey. 2024. Sparse autoencoders find highly interpretable features in language models. In The Twelfth International Conference on Learning Representations. find
  22. Yong Dai, Xiaopeng Hong, Yabin Wang, Zhiheng Ma, Jinfeng Yang, Dongmei Jiang, and Yaowei Wang. 2025. Dual-attention based prompt generation and catalyzing for instance-wise continual learning. Pattern Recognition, page 112685. find
  23. DeepSeek-AI, Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, and 181 others. 2025. Deepseek-v3 technical report. Preprint, arXiv:2… find
  24. Boyi Deng, Yu Wan, Baosong Yang, Yidan Zhang, and Fuli Feng. 2025. Unveiling language-specific features in large language models via sparse autoencoders. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4563–4608, Vienna, Austr…
  25. Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, pages 248–255. find
  26. Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, and Luke Zettlemoyer. 2023. Qlora: Efficient finetuning of quantized llms. Preprint, arXiv:2305.14314.
  27. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. Preprint, arXiv:1810.04805.
  28. Ning Ding, Yujia Qin, Guang Yang, Fuchao Wei, Zonghan Yang, Yusheng Su, Shengding Hu, Yulin Chen, Chi-Min Chan, Weize Chen, and 1 others. 2023. Parameter-efficient fine-tuning of large-scale pre-trained language models. Nature machine intelligence, 5(3):220–235.
  29. Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Jingyuan Ma, Rui Li, Heming Xia, Jingjing Xu, Zhiyong Wu, Baobao Chang, Xu Sun, Lei Li, and Zhifang Sui. 2024. A survey on in-context learning. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 1107–1128, Miami…
  30. Nelson Elhage, Neel Nanda, Catherine Olsson, Tom Henighan, Nicholas Joseph, Ben Mann, Amanda Askell, Yuntao Bai, Anna Chen, Tom Conerly, and 1 others. 2021. A mathematical framework for transformer circuits. Transformer Circuits Thread Blogpost. find
  31. Wenfeng Feng, Chuzhan Hao, Yuewei Zhang, Yu Han, and Hao Wang. 2024. Mixture-of-LoRAs: An efficient multitask tuning method for large language models. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 1… find
  32. Heshan Fernando, Han Shen, Parikshit Ram, Yi Zhou, Horst Samulowitz, Nathalie Baracaldo, and Tianyi Chen. 2024. Understanding forgetting in llm supervised fine-tuning and preference learning–a convex optimization perspective. arXiv preprint arXiv:2410.15483. — [arXiv:2410.15483](https://arxiv.org/abs/2410.15483) find
  33. Robert M French. 1999. Catastrophic forgetting in connectionist networks. Trends in cognitive sciences, 3(4):128–135. find
  34. Leo Gao, Tom Dupré la Tour, Henk Tillman, Gabriel Goh, Rajan Troll, Alec Radford, Ilya Sutskever, Jan Leike, and Jeffrey Wu. 2024. Scaling and evaluating sparse autoencoders. ArXiv:2406.04093. find
  35. Leo Gao, Achyuta Rajaram, Jacob Coxon, Soham V Govande, Bowen Baker, and Dan Mossing. 2025. Weight-sparse transformers have interpretable circuits. arXiv preprint arXiv:2511.13653. — [arXiv:2511.13653](https://arxiv.org/abs/2511.13653) find
  36. Atticus Geiger, Duligur Ibeling, Amir Zur, Maheep Chaudhary, Sonakshi Chauhan, Jing Huang, Aryaman Arora, Zhengxuan Wu, Noah Goodman, Christopher Potts, and Thomas Icard. 2024. Causal abstraction: A theoretical foundation for mechanistic interpretability. arXiv:2301.04709. find
  37. Atticus Geiger, Hanson Lu, Thomas Icard, and Christopher Potts. 2021. Causal abstractions of neural networks. In Advances in Neural Information Processing Systems, volume 34, pages 9574–9586. Curran Associates, Inc. find
  38. Jiahui Geng, Qing Li, Herbert Woisetschlaeger, Zongxiong Chen, Fengyu Cai, Yuxia Wang, Preslav Nakov, Hans-Arno Jacobsen, and Fakhri Karray. 2025. A comprehensive survey of machine unlearning techniques for large language models. arXiv preprint arXiv:2503.01854. — [arXiv:2503.01854](https://arxiv.org/abs/2503.01854) find
  39. Ross Girshick, Jeff Donahue, Trevor Darrell, and Jitendra Malik. 2014. Rich feature hierarchies for accurate object detection and semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
  40. Ian J Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2014. Generative adversarial nets. Advances in neural information processing systems, 27. find
  41. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, Amy Yang, Angela Fan, Anirudh Goyal, Anthony Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, and 542 o… — The Llama 3 Herd of Models
  42. Xu Guo, Boyang Li, and Han Yu. 2022. Improving the sample efficiency of prompt tuning with domain adaptation. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 3523– 3537, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics. find
  43. Daniil Gurgurov, Yusser Al Ghussin, Tanja Baeumel, Cheng-Ting Chou, Patrick Schramowski, Marius Mosbach, Josef van Genabith, and Simon Ostermann. 2026. Clas-bench: A cross-lingual alignment and steering benchmark. Findings of the Association for Computational Linguistics: ACL 2026. find
  44. Daniil Gurgurov, Katharina Trinley, Yusser Al Ghussin, Tanja Baeumel, Josef Van Genabith, and Simon Ostermann. 2025. Language arithmetics: Towards systematic language neuron identification and manipulation. In Proceedings of the 14th International Joint Conference on Natural Language Processing and… find
  45. Zeyu Han, Chao Gao, Jinyang Liu, Jeff Zhang, and Sai Qian Zhang. 2024. Parameter-efficient fine-tuning for large models: A comprehensive survey. Preprint, arXiv:2403.14608.
  46. Zhengfu He, Wentao Shu, Xuyang Ge, Lingjie Chen, Junxuan Wang, Yunhua Zhou, Frances Liu, Qipeng Guo, Xuanjing Huang, Zuxuan Wu, and 1 others. 2024. Llama scope: Extracting millions of features from llama-3.1-8b with sparse autoencoders. arXiv preprint arXiv:2410.20526. find
  47. Roee Hendel, Mor Geva, and Amir Globerson. 2023. In-context learning creates task vectors. arXiv preprint arXiv:2310.15916.
  48. Neil Houlsby, Andrei Giurgiu, Stanislaw Jastrzebski, Bruna Morrone, Quentin de Laroussilhe, Andrea Gesmundo, Mona Attariyan, and Sylvain Gelly. 2019. Parameter-efficient transfer learning for nlp. Preprint, arXiv:1902.00751. find
  49. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2021. Lora: Low-rank adaptation of large language models. Preprint, arXiv:2106.09685.
  50. Robert Huben, Hoagy Cunningham, Logan Riggs Smith, Aidan Ewart, and Lee Sharkey. 2024. Sparse autoencoders find highly interpretable features in language models. In The Twelfth International Conference on Learning Representations. find
  51. Gabriel Ilharco, Marco Tulio Ribeiro, Mitchell Wortsman, Suchin Gururangan, Ludwig Schmidt, Hannaneh Hajishirzi, and Ali Farhadi. 2022. Editing models with task arithmetic. arXiv preprint arXiv:2212.04089.
  52. Gabriel Ilharco, Marco Tulio Ribeiro, Mitchell Wortsman, Ludwig Schmidt, Hannaneh Hajishirzi, and Ali Farhadi. 2023. Editing models with task arithmetic. In The Eleventh International Conference on Learning Representations.
  53. Anna A. Ivanova, John Hewitt, and Noga Zaslavsky. 2021. Probing artificial neural networks: insights from neuroscience. Preprint, arXiv:2104.08197. — [arXiv:2104.08197](https://arxiv.org/abs/2104.08197) find
  54. Kai Konen, Sophie Jentzsch, Diaoulé Diallo, Peer Schütt, Oliver Bensch, Roxanne El Baff, Dominik Opitz, and Tobias Hecking. 2024. Style vectors for steering generative large language model. Preprint, arXiv:2402.01618. — [arXiv:2402.01618](https://arxiv.org/abs/2402.01618) find
  55. Suhas Kotha, Jacob Mitchell Springer, and Aditi Raghunathan. 2024. Understanding catastrophic forgetting in language models via implicit inference. In The Twelfth International Conference on Learning Representations.
  56. Brian Lester, Rami Al-Rfou, and Noah Constant. 2021. The power of scale for parameter-efficient prompt tuning. Preprint, arXiv:2104.08691.
  57. Hongyu Li, Liang Ding, Meng Fang, and Dacheng Tao. 2024. Revisiting catastrophic forgetting in large language model tuning. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 4297–4308, Miami, Florida, USA. Association for Computational Linguistics. find
  58. Kenneth Li, Oam Patel, Fernanda Viégas, Hanspeter Pfister, and Martin Wattenberg. 2023. Inference-time intervention: Eliciting truthful answers from a language model. Advances in Neural Information Processing Systems, 36:41451–41530.
  59. Xiang Lisa Li and Percy Liang. 2021. Prefix-tuning: Optimizing continuous prompts for generation. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 4582–… find
  60. Zhuowei Li, Haizhou Shi, Yunhe Gao, Di Liu, Zhenting Wang, Yuxiao Chen, Ting Liu, Long Zhao, Hao Wang, and Dimitris N Metaxas. 2025. The hidden life of tokens: Reducing hallucination of large vision-language models via visual information steering. arXiv preprint arXiv:2502.03628. — [arXiv:2502.03628](https://arxiv.org/abs/2502.03628) find
  61. Vladislav Lialin, Vijeta Deshpande, and Anna Rumshisky. 2023. Scaling down to scale up: A guide to parameter-efficient fine-tuning. arXiv preprint arXiv:2303.15647.
  62. Tom Lieberum, Senthooran Rajamanoharan, Arthur Conmy, Lewis Smith, Nicolas Sonnerat, Vikrant Varma, Janos Kramar, Anca Dragan, Rohin Shah, and Neel Nanda. 2024. Gemma scope: Open sparse autoencoders everywhere all at once on gemma 2. In Proceedings of the 7th BlackboxNLP Workshop: Analyzing and Inte… find
  63. Haokun Liu, Derek Tam, Mohammed Muqeeth, Jay Mohta, Tenghao Huang, Mohit Bansal, and Colin A Raffel. 2022. Few-shot parameter-efficient fine-tuning is better and cheaper than in-context learning. Advances in Neural Information Processing Systems, 35:1950–1965.
  64. Sheng Liu, Haotian Ye, Lei Xing, and James Y. Zou. 2024. In-context vectors: Making in context learning more effective and controllable through latent space steering. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna, Austria, July 21-27, 2024. OpenReview.net. find
  65. Sijia Liu, Yuanshun Yao, Jinghan Jia, Stephen Casper, Nathalie Baracaldo, Peter Hase, Yuguang Yao, Chris Yuhao Liu, Xiaojun Xu, Hang Li, and 1 others. 2025. Rethinking machine unlearning for large language models. Nature Machine Intelligence, pages 1–14. find
  66. Wei Lu, Rachel K Luu, and Markus J Buehler. 2025. Fine-tuning large language models for domain adaptation: Exploration of training strategies, scaling, model merging and synergistic capabilities. npj Computational Materials, 11(1):84. find
  67. Yun Luo, Zhen Yang, Fandong Meng, Yafu Li, Jie Zhou, and Yue Zhang. 2025. An empirical study of catastrophic forgetting in large language models during continual fine-tuning. IEEE Transactions on Audio, Speech and Language Processing. find
  68. Sourab Mangrulkar, Sylvain Gugger, Lysandre Debut, Younes Belkada, Sayak Paul, Benjamin Bossan, and Marian Tietz. 2022. PEFT: State-of-the-art parameter-efficient fine-tuning methods. https://github.com/huggingface/peft. ↗
  69. Yishay Mansour, Mehryar Mohri, and Afshin Rostamizadeh. 2023. Domain adaptation: Learning bounds and algorithms. Preprint, arXiv:0902.3430. — [arXiv:0902.3430](https://arxiv.org/abs/0902.3430) find
  70. Samuel Marks and Max Tegmark. 2023. The geometry of truth: Emergent linear structure in large language model representations of true/false datasets. Preprint, arXiv:2310.06824.
  71. Harry Mayne, Yushi Yang, and Adam Mahdi. 2024. Can sparse autoencoders be used to decompose and interpret steering vectors? arXiv preprint arXiv:2411.08790. — [arXiv:2411.08790](https://arxiv.org/abs/2411.08790) find
  72. Kevin Meng, David Bau, Alex Andonian, and Yonatan Belinkov. 2022a. Locating and editing factual associations in gpt. In Advances in Neural Information Processing Systems, volume 35, pages 17359–17372. Curran Associates, Inc. find
  73. Kevin Meng, David Bau, Alex Andonian, and Yonatan Belinkov. 2022b. Locating and editing factual associations in GPT. In Advances in Neural Information Processing Systems, volume 35, pages 17359–17372. Curran Associates, Inc. find
  74. Kevin Meng, David Bau, Alex Andonian, and Yonatan Belinkov. 2022c. Locating and editing factual associations in gpt. In Advances in Neural Information Processing Systems, volume 35, page 17359–17372. find
  75. Tomáš Mikolov, Wen-tau Yih, and Geoffrey Zweig. 2013. Linguistic regularities in continuous space word representations. In Proceedings of the 2013 conference of the north american chapter of the association for computational linguistics: Human language technologies, pages 746–751.
  76. Moran Mizrahi, Guy Kaplan, Dan Malkin, Rotem Dror, Dafna Shahaf, and Gabriel Stanovsky. 2024a. State of what art? a call for multi-prompt llm evaluation. Transactions of the Association for Computational Linguistics, 12:933–949. find
  77. Moran Mizrahi, Guy Kaplan, Dan Malkin, Rotem Dror, Dafna Shahaf, and Gabriel Stanovsky. 2024b. State of what art? a call for multi-prompt llm evaluation. Transactions of the Association for Computational Linguistics, 12:933–949. find
  78. Neel Nanda, Andrew Lee, and Martin Wattenberg. 2023. Emergent linear representations in world models of self-supervised sequence models. In Proceedings of the 6th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP, pages 16–30, Singapore. Association for Computational Linguisti… find
  79. Duy Nguyen, Archiki Prasad, Elias Stengel-Eskin, and Mohit Bansal. 2025. Multi-attribute steering of language models via targeted intervention. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 20619–20634, Vienna, Austria. Asso…
  80. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, and 1 others. 2022. Training language models to follow instructions with human feedback. Advances in neural information processing systems, 35:27730–27744.
  81. Nina Panickssery, Nick Gabrieli, Julian Schulz, Meg Tong, Evan Hubinger, and Alexander Matt Turner. 2023. Steering llama 2 via contrastive activation addition. arXiv preprint arXiv:2312.06681. find
  82. Kiho Park, Yo Joong Choe, and Victor Veitch. 2024. The linear representation hypothesis and the geometry of large language models. Preprint, arXiv:2311.03658.
  83. David Patterson, Joseph Gonzalez, Quoc Le, Chen Liang, Lluis-Miquel Munguia, Daniel Rothchild, David So, Maud Texier, and Jeff Dean. 2021. Carbon emissions and large neural network training. Preprint, arXiv:2104.10350.
  84. Branislav Pecher, Ivan Srba, and Maria Bielikova. 2025. Comparing specialised small and general large language models on text classification: 100 labelled samples to achieve break-Even performance. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 165–1…
  85. Jeffrey Pennington, Richard Socher, and Christopher D Manning. 2014. Glove: Global vectors for word representation. In Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP), pages 1532–1543.
  86. Matthew E. Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. 2018. Deep contextualized word representations. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Techno…
  87. Jonas Pfeiffer, Aishwarya Kamath, Andreas Rücklé, Kyunghyun Cho, and Iryna Gurevych. 2021. find
  88. Adapterfusion: Non-destructive task composition for transfer learning. In Proceedings of the 16th conference of the European chapter of the association for computational linguistics: main volume, pages 487–503. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Ming… — DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
  89. Jonas Pfeiffer, Ivan Vulić, Iryna Gurevych, and Sebastian Ruder. 2020. MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 7654–7673, Online. Association for Computational L…
  90. Viacheslav Sinii, Nikita Balagansky, Yaroslav Aksenov, Vadim Kurochkin, Daniil Laptev, Gleb Gerasimov, Alexey Gorbatovski, Boris Shaposhnikov, and Daniil Gavrilov. 2025. Small vectors, big effects: A mechanistic study of rl-induced reasoning via steering vectors. arXiv preprint arXiv:2509.06608. — [arXiv:2509.06608](https://arxiv.org/abs/2509.06608) find
  91. Joris Postmus and Steven Abreu. 2024. Steering large language models using conceptors: Improving addition-based activation engineering. In MINT: Foundation Model Interventions. find
  92. Alec Radford, Rafal Jozefowicz, and Ilya Sutskever. 2017. Learning to generate reviews and discovering sentiment. Preprint, arXiv:1704.01444.
  93. Alec Radford, Luke Metz, and Soumith Chintala. 2015. Unsupervised representation learning with deep convolutional generative adversarial networks. CoRR, abs/1511.06434.
  94. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, and 1 others. 2019. Language models are unsupervised multitask learners. OpenAI blog, 1(8):9. find
  95. Jonas B Raedler, Weiyue Li, Alyssa Mia Taliotis, Manasvi Goyal, Siddharth Swaroop, and Weiwei Pan. 2025. The necessity for intervention fidelity: Unintended side effects when steering LLMs. In ICML 2025 Workshop on Methods and Opportunities at Small Scale. find
  96. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. 2023. Direct preference optimization: Your language model is secretly a reward model. Advances in neural information processing systems, 36:53728–53741.
  97. Nina Rimsky, Nick Gabrieli, Julian Schulz, Meg Tong, Evan Hubinger, and Alexander Turner. 2024. Steering llama 2 via contrastive activation addition. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 15504–15522, Bangkok, Thaila… find
  98. Naomi Saphra and Sarah Wiegreffe. 2024. Mechanistic? Preprint, arXiv:2410.09087. — [arXiv:2410.09087](https://arxiv.org/abs/2410.09087) find
  99. John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. 2017. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347. find
  100. Pierre Sermanet, David Eigen, Xiang Zhang, Michael Mathieu, Rob Fergus, and Yann LeCun. 2014. Overfeat: Integrated recognition, localization and detection using convolutional networks. Preprint, arXiv:1312.6229. find
  101. Vincent Siu, Nicholas Crispino, David Park, Nathan W Henry, Zhun Wang, Yang Liu, Dawn Song, and Chenguang Wang. 2025. Steeringsafety: A systematic safety evaluation framework of representation steering in llms. arXiv preprint arXiv:2509.13450. — [arXiv:2509.13450](https://arxiv.org/abs/2509.13450) find
  102. Anushka Sivakumar, Andrew Zhang, Zaber Hakim, and Chris Thomas. 2025. Steervlm: Robust model control through lightweight activation steering for vision language models. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 23640–23665. find
  103. Asa Cooper Stickland, Alexander Lyzhov, Jacob Pfau, Salsabila Mahdi, and Samuel R Bowman. 2024. Steering without side effects: Improving post-deployment control of language models. arXiv preprint arXiv:2406.15518. find
  104. Jingran Su, Jingfan Chen, Hongxin Li, Yuntao Chen, Li Qing, and Zhaoxiang Zhang. 2025. Activation steering decoding: Mitigating hallucination in large vision-language models through bidirectional hidden state intervention. In Proceedings of the 63rd Annual Meeting of the Association for Computationa…
  105. Nishant Subramani, Nivedita Suresh, and Matthew E Peters. 2022. Extracting latent steering vectors from pretrained language models. arXiv preprint arXiv:2205.05124.
  106. Denis Sutter, Julian Minder, Thomas Hofmann, and Tiago Pimentel. 2025. The non-linear representation dilemma: Is causal abstraction enough for mechanistic interpretability? arXiv preprint arXiv:2507.08802. — [arXiv:2507.08802](https://arxiv.org/abs/2507.08802) find
  107. Daniel Chee Hian Tan, David Chanin, Aengus Lynch, Brooks Paige, Dimitrios Kanoulas, Adrià Garriga-Alonso, and Robert Kirk. 2024. Analysing the generalisation and reliability of steering vectors. In The Thirty-eighth Annual Conference on Neural Information Processing Systems. find
  108. Tianyi Tang, Wenyang Luo, Haoyang Huang, Dongdong Zhang, Xiaolei Wang, Wayne Xin Zhao, Furu Wei, and Ji-Rong Wen. 2024. Language-specific neurons: The key to multilingual capabilities in large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics…
  109. Adly Templeton, Tom Conerly, Jonathan Marcus, Jack Lindsey, Trenton Bricken, Brian Chen, Adam Pearce, Craig Citro, Emmanuel Ameisen, Andy Jones, and 1 others. 2024. Scaling monosemanticity: Extracting interpretable features from claude 3 sonnet. Transformer Circuits Thread Blogpost. find
  110. Eric Todd, Millicent L. Li, Arnab Sen Sharma, Aaron Mueller, Byron C. Wallace, and David Bau. 2024. Function vectors in large language models. Preprint, arXiv:2310.15213. find
  111. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, Wenyin Fu, and 49 others… — Llama 2: Open Foundation and Fine-Tuned Chat Models
  112. Alexander Matt Turner, Lisa Thiergart, Gavin Leech, David Udell, Juan J. Vazquez, Ulisse Mini, and Monte MacDiarmid. 2023. Steering language models with activation engineering. Preprint, arXiv:2308.10248.
  113. Jesse Vig, Sebastian Gehrmann, Yonatan Belinkov, Sharon Qian, Daniel Nevo, Yaron Singer, and Stuart Shieber. 2020. Investigating gender bias in language models using causal mediation analysis. In Advances in Neural Information Processing Systems, volume 33, pages 12388–12401. Curran Associates, Inc. find
  114. Zhongwei Wan, Xin Wang, Che Liu, Samiul Alam, Yu Zheng, Jiachen Liu, Zhongnan Qu, Shen Yan, Yi Zhu, Quanlu Zhang, Mosharaf Chowdhury, and Mi Zhang. 2024. Efficient large language models: A survey. Preprint, arXiv:2312.03863.
  115. Haowen Wang, Tao Sun, Congyun Jin, Yingbo Wang, Yibo Fan, Yunqi Xu, Yuliang Du, and Cong Fan. 2024a. Customizable combination of parameter-efficient modules for multi-task learning. In The Twelfth International Conference on Learning Representations. find
  116. Luping Wang, Sheng Chen, Linnan Jiang, Shu Pan, Runze Cai, Sen Yang, and Fei Yang. 2025a. Parameter-efficient fine-tuning in large language models: a survey of methodologies. Artificial Intelligence Review, 58(8):227. find
  117. Pengyu Wang, Dong Zhang, Linyang Li, Chenkun Tan, Xinghao Wang, Mozhi Zhang, Ke Ren, Botian Jiang, and Xipeng Qiu. 2024b. Inferaligner: Inference-time alignment for harmlessness through cross-model guidance. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, p… find
  118. Weixuan Wang, Minghao Wu, Barry Haddow, and Alexandra Birch. 2025b. Expertsteer: Intervening in llms through expert knowledge. arXiv preprint arXiv:2505.12313. — [arXiv:2505.12313](https://arxiv.org/abs/2505.12313) find
  119. Xinpeng Wang, Mingyang Wang, Yihong Liu, Hinrich Schuetze, and Barbara Plank. 2025c. Refusal direction is universal across safety-aligned languages. In The Thirty-ninth Annual Conference on Neural Information Processing Systems. find
  120. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, and Denny Zhou. 2023. Chain-of-thought prompting elicits reasoning in large language models. Preprint, arXiv:2201.11903.
  121. Zhengxuan Wu, Aryaman Arora, Atticus Geiger, Zheng Wang, Jing Huang, Dan Jurafsky, Christopher D Manning, and Christopher Potts. 2025. Axbench: Steering llms? even simple baselines outperform sparse autoencoders. In Forty-second International Conference on Machine Learning. find
  122. Zhengxuan Wu, Aryaman Arora, Zheng Wang, Atticus Geiger, Dan Jurafsky, Christopher D. Manning, and Christopher Potts. 2024. ReFT: Representation finetuning for language models. arXiv:2404.03592. find
  123. Yuanshun Yao, Xiaojun Xu, and Yang Liu. 2024. Large language model unlearning. Advances in Neural Information Processing Systems, 37:105425–105475.
  124. J.D. Zamfirescu-Pereira, Heather Wei, Amy Xiao, Kitty Gu, Grace Jung, Matthew G Lee, Bjoern Hartmann, and Qian Yang. 2023a. Herding ai cats: Lessons from designing a chatbot by prompting gpt-3. In Proceedings of the 2023 ACM Designing Interactive Systems Conference, DIS ’23, page 2206–2220, New York… find
  125. J.D. Zamfirescu-Pereira, Richmond Y. Wong, Bjoern Hartmann, and Qian Yang. 2023b. Why johnny can’t prompt: How non-ai experts try (and fail) to design llm prompts. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, CHI ’23, New York, NY, USA. Association for Computing M… find
  126. Shengyu Zhang, Linfeng Dong, Xiaoya Li, Sen Zhang, Xiaofei Sun, Shuhe Wang, Jiwei Li, Runyi Hu, Tianwei Zhang, Fei Wu, and Guoyin Wang. 2025. Instruction tuning for large language models: A survey. Preprint, arXiv:2308.10792.
  127. Haiyan Zhao, Xuansheng Wu, Fan Yang, Bo Shen, Ninghao Liu, and Mengnan Du. 2026. Denoising concept vectors with sparse autoencoders for improved language model steering. In Findings of the Association for Computational Linguistics: EACL 2026, pages 797–808, Rabat, Morocco. Association for Computatio…
  128. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, Yifan Du, Chen Yang, Yushuo Chen, Zhipeng Chen, Jinhao Jiang, Ruiyang Ren, Yifan Li, Xinyu Tang, Zikang Liu, and 3 others. 2025a. A survey of large language models. Prepr…
  129. Yiran Zhao, Wenxuan Zhang, Guizhen Chen, Kenji Kawaguchi, and Lidong Bing. 2024. How do large language models handle multilingualism? Advances in Neural Information Processing Systems, 37:15296– 15319.
  130. Ziyu Zhao, Tao Shen, Didi Zhu, Zexi Li, Jing Su, Xuwu Wang, and Fei Wu. 2025b. Merging loRAs like playing LEGO: Pushing the modularity of loRA to extremes through rank-wise clustering. In The Thirteenth International Conference on Learning Representations. find
  131. Yuxiang Zhou, Jiazheng Li, Yanzheng Xiang, Hanqi Yan, Lin Gui, and Yulan He. 2024. The mystery of in-context learning: A comprehensive survey on interpretation and analysis. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 14365–14378, Miami, Florida,… find
  132. Andy Zou, Long Phan, Sarah Chen, James Campbell, Phillip Guo, Richard Ren, Alexander Pan, Xuwang Yin, Mantas Mazeika, Ann-Kathrin Dombrowski, Shashwat Goel, Nathaniel Li, Michael J. Byun, Zifan Wang, Alex Mallen, Steven Basart, Sanmi Koyejo, Dawn Song, Matt Fredrikson, and 2 others. 2023. Representa…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/ostermann-2026-weights-activations-is-steering-next/