{"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-2607-00605","title":"Auditing Forgetting in Limited Memory Language Models","arxiv_id":"2607.00605","date":"2026-07-01","proceeding":null,"authors":["Arya Raeesi","Hanna Roed"],"abstract":"Limited Memory Language Models (LMLMs) externalize factual knowledge to a database to enable deletion-based unlearning without retraining. Existing evaluations measure post-deletion correctness in aggregate and cannot tell whether a deleted fact persists through residual parametric memory, alternative retrieval paths, or near-neighbor retrieval artifacts. We propose a causal auditing framework that holds the model fixed and varies the database state at inference time across three interventions: FULL, DEL-ON, and DEL-OFF. The framework decomposes post-deletion behavior into parametric leakage L(f), retrieval-mediated correctness R(f), and a retrieval artifact rate grounded in the inference-time retrieval trace. We apply it to 12,228 alias-closure deletions across thirteen databases, including four adversarial topologies (Base, Alias, Noise, Collision) we construct in three domains, and six prompt formulations. Parametric leakage is near zero in every variant and every prompt style: the model rarely returns the deleted answer in the absence of retrieval. The residual that does survive lives in the retrieval graph: retrieval-mediated correctness and the retrieval artifact rate match within rounding everywhere, so post-deletion correctness is, in our audit, predominantly reconstituted from near-neighbor retrieval. This residual ranges from 0.7% on the released LMLM database to 13.6% on the most adversarial variant, and prompt formulation does not independently control how much of a deleted fact survives. These results suggest that, for this class of LMLM and deletion procedure, the unlearning boundary is drawn primarily by the database administrator rather than by the model.","url_abs":"https://arxiv.org/abs/2607.00605","url_pdf":"https://arxiv.org/pdf/2607.00605","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":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2607.00605","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2607.00605"}},"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/raeesiarya/LMLMAudit","reach":null}],"summary":{"ran_draft_wrong":12,"ran_violates":4,"ran":4,"unverified":6},"by_repo_kind":{"found_in_text":{"samples":26,"ran":20,"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":"bf84ec4bafaa142c","entry":"_default_retrieval_trace","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"bf84ec4bafaa142c"}},{"code_sha256_prefix":"9fd2a6d4256cac2e","entry":"_flatten_alias_values","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9fd2a6d4256cac2e"}},{"code_sha256_prefix":"d75287169d8f8eff","entry":"_unique_preserving_order","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d75287169d8f8eff"}},{"code_sha256_prefix":"10a3c23b4e022f14","entry":"build_alias_set","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"10a3c23b4e022f14"}},{"code_sha256_prefix":"b240caf9e4324995","entry":"candidate_supports_target_fact","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b240caf9e4324995"}},{"code_sha256_prefix":"2625b513a8784327","entry":"choose_answer","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2625b513a8784327"}},{"code_sha256_prefix":"e78ccae7868d860d","entry":"clean_answer","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e78ccae7868d860d"}},{"code_sha256_prefix":"ab09f379e654b1e5","entry":"compute_generation_budget","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ab09f379e654b1e5"}},{"code_sha256_prefix":"0373bc411296aa7d","entry":"extract_lookup_query","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0373bc411296aa7d"}},{"code_sha256_prefix":"9c4e338ec3112dee","entry":"extract_lookup_values","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9c4e338ec3112dee"}},{"code_sha256_prefix":"656b381216db2a07","entry":"is_deleted_triplet","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"656b381216db2a07"}},{"code_sha256_prefix":"c1f14f57e9136ea6","entry":"normalize_text","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c1f14f57e9136ea6"}},{"code_sha256_prefix":"f7a44c18a1cf3b39","entry":"prepare_prompt","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f7a44c18a1cf3b39"}},{"code_sha256_prefix":"23910038515b997b","entry":"prompt_row_aliases","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"23910038515b997b"}},{"code_sha256_prefix":"09d4a1065936f6a3","entry":"retrieval_enabled","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"09d4a1065936f6a3"}},{"code_sha256_prefix":"c5f599a0ba29237c","entry":"retrieve_lookup_value","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c5f599a0ba29237c"}},{"code_sha256_prefix":"3daeab8516d2d741","entry":"retrieve_triplet_candidates","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3daeab8516d2d741"}},{"code_sha256_prefix":"65fdf054022cc6b8","entry":"target_fact_from_prompt_row","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"65fdf054022cc6b8"}},{"code_sha256_prefix":"56c931802f771660","entry":"tokenize","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"56c931802f771660"}},{"code_sha256_prefix":"6d58c6065b37de13","entry":"values_equivalent","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6d58c6065b37de13"}},{"code_sha256_prefix":"6ae72e492d22758d","entry":"AuditDatabaseManager","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6ae72e492d22758d"}},{"code_sha256_prefix":"5372af2c232205bf","entry":"DatabaseState","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5372af2c232205bf"}},{"code_sha256_prefix":"bef54e1f3b6ad335","entry":"TargetFact","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"bef54e1f3b6ad335"}},{"code_sha256_prefix":"a3789f0f9efa93ad","entry":"build_state_db_manager","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a3789f0f9efa93ad"}},{"code_sha256_prefix":"1908bdfde38728bb","entry":"generate_answer","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1908bdfde38728bb"}},{"code_sha256_prefix":"69563b95bc69fdff","entry":"run_prompt_audit","repo":"raeesiarya/LMLMAudit","repo_kind":"found_in_text","path":"src/lmlm-audit/run_audit.py","file_url":"https://github.com/raeesiarya/LMLMAudit/blob/HEAD/src/lmlm-audit/run_audit.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"69563b95bc69fdff"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":{"kind":"leaderboard_placements","source":"Syntology's leaderboard-shaped extractor over the paper's own arXiv-HTML tables: a model pointed at a cell, the number was read from that cell and checked against the board's metric, dataset, split and scale, and an independent check accepted the entry; not reviewed by the paper's authors or the archive's editors","extractor_model":"global.anthropic.claude-sonnet-4-5-20250929-v1:0","verifier_model":null,"prompt_sha":"fa63d4bb9d755694","coverage":{"sentence":"Syntology has checked 6,795 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","papers_newer_than_archive":9581,"papers_checked":6795},"entries":[],"not_placed":{"boards":0,"rejected_by_independent_check":0,"refused_by_a_rule":0,"check_did_not_answer":0,"proposed_without_a_cell":0,"declined_by_site":0}}}