{"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/does-representation-intervention-really","title":"Does Representation Intervention Really Identify Desired Concepts and Elicit Alignment?","arxiv_id":"2505.18672","date":"2025-05-24","proceeding":null,"authors":["Hongzheng Yang","Yongqiang Chen","Zeyu Qin","Tongliang Liu","Chaowei Xiao","Kun Zhang","Bo Han"],"abstract":"Representation intervention aims to locate and modify the representations that encode the underlying concepts in Large Language Models (LLMs) to elicit the aligned and expected behaviors. Despite the empirical success, it has never been examined whether one could locate the faithful concepts for intervention. In this work, we explore the question in safety alignment. If the interventions are faithful, the intervened LLMs should erase the harmful concepts and be robust to both in-distribution adversarial prompts and the out-of-distribution (OOD) jailbreaks. While it is feasible to erase harmful concepts without degrading the benign functionalities of LLMs in linear settings, we show that it is infeasible in the general non-linear setting. To tackle the issue, we propose Concept Concentration (COCA). Instead of identifying the faithful locations to intervene, COCA refractors the training data with an explicit reasoning process, which firstly identifies the potential unsafe concepts and then decides the responses. Essentially, COCA simplifies the decision boundary between harmful and benign representations, enabling more effective linear erasure. Extensive experiments with multiple representation intervention methods and model architectures demonstrate that COCA significantly reduces both in-distribution and OOD jailbreak success rates, and meanwhile maintaining strong performance on regular tasks such as math and code generation.","url_abs":"https://arxiv.org/abs/2505.18672v1","url_pdf":"https://arxiv.org/pdf/2505.18672v1.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":[],"tasks":[{"task_slug":"code-generation","task_name":"Code Generation"},{"task_slug":"math","task_name":"Math"},{"task_slug":"safety-alignment","task_name":"Safety Alignment"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2505.18672","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.18672"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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":"deterministic:regex_extraction","url":"https://github.com/tmlr-group/COCA","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran":2,"unverified":2},"by_repo_kind":{"found_in_text":{"samples":4,"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":0,"samples":[{"code_sha256_prefix":"dbea5358be6b5ad0","entry":"create_score_evaluation_response","repo":"tmlr-group/COCA","repo_kind":"found_in_text","path":"src/llamafactory/api/chat.py","file_url":"https://github.com/tmlr-group/COCA/blob/HEAD/src/llamafactory/api/chat.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"dbea5358be6b5ad0"}},{"code_sha256_prefix":"57583cf880d95444","entry":"jsonify","repo":"tmlr-group/COCA","repo_kind":"found_in_text","path":"src/llamafactory/api/common.py","file_url":"https://github.com/tmlr-group/COCA/blob/HEAD/src/llamafactory/api/common.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"57583cf880d95444"}},{"code_sha256_prefix":"87aa9acaa81294f5","entry":"dictify","repo":"tmlr-group/COCA","repo_kind":"found_in_text","path":"src/llamafactory/api/common.py","file_url":"https://github.com/tmlr-group/COCA/blob/HEAD/src/llamafactory/api/common.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"87aa9acaa81294f5"}},{"code_sha256_prefix":"14abb4615516bc7e","entry":"get_eval_template","repo":"tmlr-group/COCA","repo_kind":"found_in_text","path":"src/llamafactory/eval/template.py","file_url":"https://github.com/tmlr-group/COCA/blob/HEAD/src/llamafactory/eval/template.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"14abb4615516bc7e"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}