{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/safety-arithmetic-a-framework-for-test-time","title":"Safety Arithmetic: A Framework for Test-time Safety Alignment of Language Models by Steering Parameters and Activations","arxiv_id":"2406.11801","date":"2024-06-17","proceeding":null,"authors":["Rima Hazra","Sayan Layek","Somnath Banerjee","Soujanya Poria"],"abstract":"Ensuring the safe alignment of large language models (LLMs) with human values is critical as they become integral to applications like translation and question answering. Current alignment methods struggle with dynamic user intentions and complex objectives, making models vulnerable to generating harmful content. We propose Safety Arithmetic, a training-free framework enhancing LLM safety across different scenarios: Base models, Supervised fine-tuned models (SFT), and Edited models. Safety Arithmetic involves Harm Direction Removal to avoid harmful content and Safety Alignment to promote safe responses. Additionally, we present NoIntentEdit, a dataset highlighting edit instances that could compromise model safety if used unintentionally. Our experiments show that Safety Arithmetic significantly improves safety measures, reduces over-safety, and maintains model utility, outperforming existing methods in ensuring safe content generation.","url_abs":"https://arxiv.org/abs/2406.11801v2","url_pdf":"https://arxiv.org/pdf/2406.11801v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"safety-arithmetic-a-framework-for-test-time","repo_url":"https://github.com/declare-lab/safety-arithmetic","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"ai-and-safety","task_name":"AI and Safety"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"safety-alignment","task_name":"Safety Alignment"}],"methods":[{"method_slug":"base","method_name":"BASE"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2406.11801","atlas_url":"https://app.syntology.ai/?focus=2406.11801","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.11801"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/declare-lab/safety-arithmetic","reach":null},{"provenance":"deterministic:regex_extraction","url":"https://github.com/hiyouga/LLaMA-Factory","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran_fixture":2},"by_repo_kind":{"official":{"samples":2,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":2,"samples":[{"code_sha256_prefix":"f34e7fb678fc843c","entry":"ties_merging","repo":"declare-lab/safety-arithmetic","repo_kind":"official","path":"HDR/Ties_Merge.py","file_url":"https://github.com/declare-lab/safety-arithmetic/blob/HEAD/HDR/Ties_Merge.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f34e7fb678fc843c"}},{"code_sha256_prefix":"3258283dc4ff9ef0","entry":"topk_values_mask","repo":"declare-lab/safety-arithmetic","repo_kind":"official","path":"HDR/Ties_Merge.py","file_url":"https://github.com/declare-lab/safety-arithmetic/blob/HEAD/HDR/Ties_Merge.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"3258283dc4ff9ef0"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}