{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,15]],"date-time":"2025-11-15T07:32:47Z","timestamp":1763191967808,"version":"3.45.0"},"reference-count":31,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,6,30]],"date-time":"2025-06-30T00:00:00Z","timestamp":1751241600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,6,30]],"date-time":"2025-06-30T00:00:00Z","timestamp":1751241600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001659","name":"Deutsche Forschungsgemeinschaft","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001659","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,6,30]]},"DOI":"10.1109\/ijcnn64981.2025.11228760","type":"proceedings-article","created":{"date-parts":[[2025,11,14]],"date-time":"2025-11-14T18:46:15Z","timestamp":1763145975000},"page":"1-7","source":"Crossref","is-referenced-by-count":0,"title":["Refusal Behavior in Large Language Models: A Nonlinear Perspective"],"prefix":"10.1109","author":[{"given":"Fabian","family":"Hildebrandt","sequence":"first","affiliation":[{"name":"FAU Erlangen-Nürnberg,CCN Group Pattern Recog. Lab.,Erlangen,Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Andreas","family":"Maier","sequence":"additional","affiliation":[{"name":"FAU Erlangen-Nürnberg,Pattern Recog. Lab.,Erlangen,Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Patrick","family":"Krauss","sequence":"additional","affiliation":[{"name":"FAU Erlangen-Nürnberg,CCN Group Pattern Recog. Lab.,Erlangen,Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Achim","family":"Schilling","sequence":"additional","affiliation":[{"name":"FAU Erlangen-Nürnberg,CCN Group Pattern Recog. Lab.,Erlangen,Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1016\/j.tics.2007.07.003"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1038\/mp.2008.104"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1016\/j.biopsych.2010.07.005"},{"key":"ref4","doi-asserted-by":"crossref","DOI":"10.23915\/distill.00024.001","article-title":"Zoom in: An introduction to circuits","volume-title":"Distill","author":"Olah","year":"2020"},{"article-title":"Toy models of superposition","year":"2022","author":"Elhage","key":"ref5"},{"key":"ref6","article-title":"Towards monosemanticity: Decomposing language models with dictionary learning","volume-title":"Transformer Circuits Thread","author":"Bricken","year":"2023"},{"key":"ref7","doi-asserted-by":"crossref","DOI":"10.3115\/v1\/P15-1144","article-title":"Sparse overcomplete word vector representations","author":"Faruqui","year":"2015"},{"article-title":"Decoding the thought vector","year":"2016","author":"Goh","key":"ref8"},{"key":"ref9","doi-asserted-by":"crossref","DOI":"10.1162\/tacl_a_00034","article-title":"Linear algebraic structure of word senses, with applications to polysemy","author":"Arora","year":"2018"},{"article-title":"Representation engineering: A top-down approach to ai transparency","year":"2023","author":"Zou","key":"ref10"},{"article-title":"Steering language models with activation engineering","year":"2024","author":"Turner","key":"ref11"},{"article-title":"Steering llama 2 via contrastive activation addition","year":"2024","author":"Panickssery","key":"ref12"},{"key":"ref13","doi-asserted-by":"crossref","DOI":"10.52202\/079017-4322","article-title":"Refusal in language models is mediated by a single direction","author":"Arditi","year":"2024"},{"article-title":"Diff-in-means concept editing is worst-case optimal: Explaining a result by sam marks and max tegmark","year":"2023","author":"Belrose","key":"ref14"},{"article-title":"Not all language model features are linear","year":"2024","author":"Engels","key":"ref15"},{"article-title":"Stanford alpaca: An instruction-following llama model","year":"2023","author":"Taori","key":"ref16"},{"article-title":"harmless_alpaca","year":"2024","author":"Labonne","key":"ref17"},{"article-title":"Universal and transferable adversarial attacks on aligned language models","year":"2023","author":"Zou","key":"ref18"},{"article-title":"harmful_behaviors","year":"2024","author":"Labonne","key":"ref19"},{"year":"2023","key":"ref20","article-title":"Model cards and prompt formats - llama 3.2"},{"article-title":"Bloom: A 176b-parameter open-access multilingual language model","year":"2023","author":"W","key":"ref21"},{"article-title":"Qwen2 technical report","year":"2024","author":"Y","key":"ref22"},{"key":"ref23","article-title":"Uncensor any llm with abliteration","volume-title":"Hugging Face Community Blog","author":"Labonne","year":"2024"},{"article-title":"Transformerlens","year":"2022","author":"Nanda","key":"ref24"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1080\/14786440109462720"},{"key":"ref26","first-page":"2579","article-title":"Visualizing data using t-SNE","volume":"9","author":"van der Maaten","year":"2008","journal-title":"Journal of Machine Learning Research"},{"article-title":"Umap: Uniform manifold approximation and projection for dimension reduction","year":"2018","author":"McInnes","key":"ref27"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2021.03.035"},{"article-title":"What features in prompts jailbreak llms? investigating the mechanisms behind attacks","year":"2024","author":"Kirch","key":"ref29"},{"article-title":"Alignment faking in large language models","year":"2024","author":"Greenblatt","key":"ref30"},{"article-title":"Refusal-llms","year":"2025","author":"Hildebrandt","key":"ref31"}],"event":{"name":"2025 International Joint Conference on Neural Networks (IJCNN)","start":{"date-parts":[[2025,6,30]]},"location":"Rome, Italy","end":{"date-parts":[[2025,7,5]]}},"container-title":["2025 International Joint Conference on Neural Networks (IJCNN)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11227166\/11227148\/11228760.pdf?arnumber=11228760","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,15]],"date-time":"2025-11-15T07:28:26Z","timestamp":1763191706000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11228760\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,30]]},"references-count":31,"URL":"https:\/\/doi.org\/10.1109\/ijcnn64981.2025.11228760","relation":{},"subject":[],"published":{"date-parts":[[2025,6,30]]}}}