{"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/arxiv-2604-17396","title":"Representation-Guided Parameter-Efficient LLM Unlearning","arxiv_id":"2604.17396","date":"2026-04-19","proceeding":null,"authors":["Zeguan Xiao","Lang Mo","Yun Chen","Lei Yang","Jiehui Zhao","Lili Yang","Guanhua Chen"],"abstract":"Large Language Models (LLMs) often memorize sensitive or harmful information, necessitating effective machine unlearning techniques. While existing parameter-efficient unlearning methods have shown promise, they still struggle with the forget-retain trade-off. This can be attributed to their reliance on parameter importance metrics to identify parameters that are important exclusively for the forget set, which is fundamentally limited by the superposition phenomenon. Due to the polysemantic nature of LLM parameters, such an importance metric may struggle to disentangle parameters associated with the forget and retain sets. In this work, we propose Representation-Guided Low-rank Unlearning (REGLU), a novel approach that leverages the geometric properties of representation spaces to achieve robust and precise unlearning. First, we develop a representation-guided initialization for LoRA that identifies the optimal subspace for selective forgetting. Second, we introduce a regularization loss that constrains the outputs of the LoRA update to lie in the orthogonal complement of the retain set's representation subspace, thereby minimizing interference with the model's performance on the retain set. We evaluate REGLU on the TOFU and WMDP benchmarks across multiple models. Our results demonstrate that REGLU consistently outperforms state-of-the-art baselines, achieving superior unlearning quality while maintaining higher model utility.","url_abs":"https://arxiv.org/abs/2604.17396","url_pdf":"https://arxiv.org/pdf/2604.17396","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2604.17396","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2604.17396"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/sustech-nlp/ReGLU","reach":null}],"summary":{"unverified":12},"by_repo_kind":{"found_in_text":{"samples":12,"ran":0,"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":12,"samples":[{"code_sha256_prefix":"5c6aaa26ba4da8f5","entry":"canonicalize_command","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/config.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/config.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"5c6aaa26ba4da8f5"}},{"code_sha256_prefix":"56a07d47b6e22052","entry":"compute_forget_quality","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/eval/tofu_aggregate.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/eval/tofu_aggregate.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"56a07d47b6e22052"}},{"code_sha256_prefix":"8bbb93f072f28c60","entry":"compute_model_utility","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/eval/tofu_aggregate.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/eval/tofu_aggregate.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"8bbb93f072f28c60"}},{"code_sha256_prefix":"f6deca5ef8ff6fd0","entry":"ensure_run_layout","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/artifacts.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/artifacts.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"f6deca5ef8ff6fd0"}},{"code_sha256_prefix":"e9e263fa340ec666","entry":"get_model_spec","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/models/registry.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/models/registry.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"e9e263fa340ec666"}},{"code_sha256_prefix":"3001fab9d6933f0f","entry":"is_peft_adapter_dir","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/trainers/common.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/trainers/common.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"3001fab9d6933f0f"}},{"code_sha256_prefix":"358c38266644a1cb","entry":"load_rol_bases","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/methods/reglu/cache.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/methods/reglu/cache.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"358c38266644a1cb"}},{"code_sha256_prefix":"1045e5778dc8f70b","entry":"main","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/validation.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/validation.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"1045e5778dc8f70b"}},{"code_sha256_prefix":"b709f503d9bc1c3e","entry":"normalize_public_command","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/config.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/config.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"b709f503d9bc1c3e"}},{"code_sha256_prefix":"02d72d5d2bb5fb16","entry":"resolve_torch_dtype","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/trainers/common.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/trainers/common.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"02d72d5d2bb5fb16"}},{"code_sha256_prefix":"30e8be9704381c2e","entry":"run_check","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/validation.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/validation.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"30e8be9704381c2e"}},{"code_sha256_prefix":"a03daa2d300b61a7","entry":"run_manifest","repo":"sustech-nlp/ReGLU","repo_kind":"found_in_text","path":"src/reglu/validation.py","file_url":"https://github.com/sustech-nlp/ReGLU/blob/HEAD/src/reglu/validation.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"a03daa2d300b61a7"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CL","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}