{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:XLYU654C4RUXH4M4S4D46ASLZ3","short_pith_number":"pith:XLYU654C","canonical_record":{"source":{"id":"2601.01972","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-05T10:27:19Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"2e0bcb19e80d0349d0fdc24f12870751dcde889c95cf9f568bd49e86f84ffe28","abstract_canon_sha256":"280a24a132efba1de55e29405797d479ae87fbabcf2744fb8d296e8250bdacee"},"schema_version":"1.0"},"canonical_sha256":"baf14f7782e46973f19c9707cf024bcedd6828d1905a8257862e7a9a3bd2fd70","source":{"kind":"arxiv","id":"2601.01972","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.01972","created_at":"2026-05-17T23:39:04Z"},{"alias_kind":"arxiv_version","alias_value":"2601.01972v4","created_at":"2026-05-17T23:39:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.01972","created_at":"2026-05-17T23:39:04Z"},{"alias_kind":"pith_short_12","alias_value":"XLYU654C4RUX","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"XLYU654C4RUXH4M4","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"XLYU654C","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:XLYU654C4RUXH4M4S4D46ASLZ3","target":"record","payload":{"canonical_record":{"source":{"id":"2601.01972","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-05T10:27:19Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"2e0bcb19e80d0349d0fdc24f12870751dcde889c95cf9f568bd49e86f84ffe28","abstract_canon_sha256":"280a24a132efba1de55e29405797d479ae87fbabcf2744fb8d296e8250bdacee"},"schema_version":"1.0"},"canonical_sha256":"baf14f7782e46973f19c9707cf024bcedd6828d1905a8257862e7a9a3bd2fd70","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:04.525096Z","signature_b64":"0tDBI8cBM6SXegtDfJhGBiKwcCR3aV8J6gMM/OVpWiBo4G4q5LBSQRTvj22YhToRkxIgPKYFa5WSCU8KAUnzAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"baf14f7782e46973f19c9707cf024bcedd6828d1905a8257862e7a9a3bd2fd70","last_reissued_at":"2026-05-17T23:39:04.524425Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:04.524425Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2601.01972","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eHVfXnmAJXGbSNo/pMCFMnUJS4YFlYQqTUgQMHivDhDM2mTP1kmUwjUH30a7iMEQ+3aC4J2EcgECkYZxhD8KBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T10:24:02.176013Z"},"content_sha256":"5025630261828b3b1181c5bfb1540da0fe92cf2722af831cf00bcb5202381aa4","schema_version":"1.0","event_id":"sha256:5025630261828b3b1181c5bfb1540da0fe92cf2722af831cf00bcb5202381aa4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:XLYU654C4RUXH4M4S4D46ASLZ3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Hidden State Poisoning Attacks against Mamba-based Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Short input phrases can irreversibly overwrite hidden states in Mamba models, inducing amnesia on retrieval tasks that pure Transformers resist.","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Alexandre Le Mercier, Chris Develder, Thomas Demeester","submitted_at":"2026-01-05T10:27:19Z","abstract_excerpt":"State space models (SSMs) like Mamba offer efficient alternatives to Transformer-based language models, with linear time complexity. Yet, their adversarial robustness remains critically unexplored. This paper studies the phenomenon whereby specific short input phrases induce a partial amnesia effect in such models, by irreversibly overwriting information in their hidden states, referred to as a Hidden State Poisoning Attack (HiSPA). Our benchmark RoBench-25 allows evaluating a model's information retrieval capabilities when subject to HiSPAs, and confirms the vulnerability of SSMs against such"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Specific short input phrases induce a partial amnesia effect in Mamba-based models by irreversibly overwriting information in their hidden states, confirmed by collapse on RoBench-25 while pure Transformers remain unaffected.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the observed performance drops are caused specifically by irreversible hidden-state overwriting rather than other mechanisms such as attention disruption or output formatting changes, and that RoBench-25 isolates this effect without confounding factors.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Short input phrases can irreversibly overwrite hidden states in Mamba models, impairing information retrieval on a new benchmark while leaving pure Transformer models unaffected.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Short input phrases can irreversibly overwrite hidden states in Mamba models, inducing amnesia on retrieval tasks that pure Transformers resist.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"b4701f3450a2a9968c79a376510e18607e1a66592cf234c6d26d6ad247c438eb"},"source":{"id":"2601.01972","kind":"arxiv","version":4},"verdict":{"id":"2002e7c8-50e3-4adc-9373-0b2a053178d9","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T18:12:20.748575Z","strongest_claim":"Specific short input phrases induce a partial amnesia effect in Mamba-based models by irreversibly overwriting information in their hidden states, confirmed by collapse on RoBench-25 while pure Transformers remain unaffected.","one_line_summary":"Short input phrases can irreversibly overwrite hidden states in Mamba models, impairing information retrieval on a new benchmark while leaving pure Transformer models unaffected.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the observed performance drops are caused specifically by irreversible hidden-state overwriting rather than other mechanisms such as attention disruption or output formatting changes, and that RoBench-25 isolates this effect without confounding factors.","pith_extraction_headline":"Short input phrases can irreversibly overwrite hidden states in Mamba models, inducing amnesia on retrieval tasks that pure Transformers resist."},"references":{"count":25,"sample":[{"doi":"","year":null,"title":"arXiv preprint arXiv:2402.01771 , year=","work_id":"884dfd62-5356-4dbe-ad95-772813d81cca","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Nemotron-h: A family of accurate and efficient hybrid mamba-transformer models","work_id":"134d0eac-bc97-4065-9c60-e8fde9d20b48","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Mahabaleshwarkar, Shih- Yang Liu, Matthijs Van Keirsbilck, Min-Hung Chen, Yoshi Suhara, et al","work_id":"283d6d50-8219-44e7-94e7-eb6b8673dec7","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Investigating the indirect object identification circuit in mamba.arXiv preprint arXiv:2407.14008,","work_id":"8d48c229-d2f2-4228-9412-6a21c701ea60","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"The Pile: An 800GB Dataset of Diverse Text for Language Modeling","work_id":"9b10667a-da61-4358-aceb-10578234d45d","ref_index":5,"cited_arxiv_id":"2101.00027","is_internal_anchor":true}],"resolved_work":25,"snapshot_sha256":"27dfefc519c323130591e8526b9e56914a3276ee7d8a3d1b2834400fa7016149","internal_anchors":5},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"2002e7c8-50e3-4adc-9373-0b2a053178d9"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EGkS5JVg7DrttdDCwV7fhiz8ftoPlKsG5eH1FrSI8fnMPTtF+9u0GCRuaP3dtCt8mF8hMiFQnaYtWvjIKP7HDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T10:24:02.176558Z"},"content_sha256":"8639f44d3bf9aa2ba5ee3f144404164d02bbb1574ad490ef3a5d7e94262e535c","schema_version":"1.0","event_id":"sha256:8639f44d3bf9aa2ba5ee3f144404164d02bbb1574ad490ef3a5d7e94262e535c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XLYU654C4RUXH4M4S4D46ASLZ3/bundle.json","state_url":"https://pith.science/pith/XLYU654C4RUXH4M4S4D46ASLZ3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XLYU654C4RUXH4M4S4D46ASLZ3/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-02T10:24:02Z","links":{"resolver":"https://pith.science/pith/XLYU654C4RUXH4M4S4D46ASLZ3","bundle":"https://pith.science/pith/XLYU654C4RUXH4M4S4D46ASLZ3/bundle.json","state":"https://pith.science/pith/XLYU654C4RUXH4M4S4D46ASLZ3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XLYU654C4RUXH4M4S4D46ASLZ3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:XLYU654C4RUXH4M4S4D46ASLZ3","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"280a24a132efba1de55e29405797d479ae87fbabcf2744fb8d296e8250bdacee","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-05T10:27:19Z","title_canon_sha256":"2e0bcb19e80d0349d0fdc24f12870751dcde889c95cf9f568bd49e86f84ffe28"},"schema_version":"1.0","source":{"id":"2601.01972","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.01972","created_at":"2026-05-17T23:39:04Z"},{"alias_kind":"arxiv_version","alias_value":"2601.01972v4","created_at":"2026-05-17T23:39:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.01972","created_at":"2026-05-17T23:39:04Z"},{"alias_kind":"pith_short_12","alias_value":"XLYU654C4RUX","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"XLYU654C4RUXH4M4","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"XLYU654C","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:8639f44d3bf9aa2ba5ee3f144404164d02bbb1574ad490ef3a5d7e94262e535c","target":"graph","created_at":"2026-05-17T23:39:04Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Specific short input phrases induce a partial amnesia effect in Mamba-based models by irreversibly overwriting information in their hidden states, confirmed by collapse on RoBench-25 while pure Transformers remain unaffected."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the observed performance drops are caused specifically by irreversible hidden-state overwriting rather than other mechanisms such as attention disruption or output formatting changes, and that RoBench-25 isolates this effect without confounding factors."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Short input phrases can irreversibly overwrite hidden states in Mamba models, impairing information retrieval on a new benchmark while leaving pure Transformer models unaffected."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Short input phrases can irreversibly overwrite hidden states in Mamba models, inducing amnesia on retrieval tasks that pure Transformers resist."}],"snapshot_sha256":"b4701f3450a2a9968c79a376510e18607e1a66592cf234c6d26d6ad247c438eb"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"State space models (SSMs) like Mamba offer efficient alternatives to Transformer-based language models, with linear time complexity. Yet, their adversarial robustness remains critically unexplored. This paper studies the phenomenon whereby specific short input phrases induce a partial amnesia effect in such models, by irreversibly overwriting information in their hidden states, referred to as a Hidden State Poisoning Attack (HiSPA). Our benchmark RoBench-25 allows evaluating a model's information retrieval capabilities when subject to HiSPAs, and confirms the vulnerability of SSMs against such","authors_text":"Alexandre Le Mercier, Chris Develder, Thomas Demeester","cross_cats":["cs.AI","cs.LG"],"headline":"Short input phrases can irreversibly overwrite hidden states in Mamba models, inducing amnesia on retrieval tasks that pure Transformers resist.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-05T10:27:19Z","title":"Hidden State Poisoning Attacks against Mamba-based Language Models"},"references":{"count":25,"internal_anchors":5,"resolved_work":25,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"arXiv preprint arXiv:2402.01771 , year=","work_id":"884dfd62-5356-4dbe-ad95-772813d81cca","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Nemotron-h: A family of accurate and efficient hybrid mamba-transformer models","work_id":"134d0eac-bc97-4065-9c60-e8fde9d20b48","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Mahabaleshwarkar, Shih- Yang Liu, Matthijs Van Keirsbilck, Min-Hung Chen, Yoshi Suhara, et al","work_id":"283d6d50-8219-44e7-94e7-eb6b8673dec7","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Investigating the indirect object identification circuit in mamba.arXiv preprint arXiv:2407.14008,","work_id":"8d48c229-d2f2-4228-9412-6a21c701ea60","year":null},{"cited_arxiv_id":"2101.00027","doi":"","is_internal_anchor":true,"ref_index":5,"title":"The Pile: An 800GB Dataset of Diverse Text for Language Modeling","work_id":"9b10667a-da61-4358-aceb-10578234d45d","year":null}],"snapshot_sha256":"27dfefc519c323130591e8526b9e56914a3276ee7d8a3d1b2834400fa7016149"},"source":{"id":"2601.01972","kind":"arxiv","version":4},"verdict":{"created_at":"2026-05-16T18:12:20.748575Z","id":"2002e7c8-50e3-4adc-9373-0b2a053178d9","model_set":{"reader":"grok-4.3"},"one_line_summary":"Short input phrases can irreversibly overwrite hidden states in Mamba models, impairing information retrieval on a new benchmark while leaving pure Transformer models unaffected.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Short input phrases can irreversibly overwrite hidden states in Mamba models, inducing amnesia on retrieval tasks that pure Transformers resist.","strongest_claim":"Specific short input phrases induce a partial amnesia effect in Mamba-based models by irreversibly overwriting information in their hidden states, confirmed by collapse on RoBench-25 while pure Transformers remain unaffected.","weakest_assumption":"That the observed performance drops are caused specifically by irreversible hidden-state overwriting rather than other mechanisms such as attention disruption or output formatting changes, and that RoBench-25 isolates this effect without confounding factors."}},"verdict_id":"2002e7c8-50e3-4adc-9373-0b2a053178d9"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:5025630261828b3b1181c5bfb1540da0fe92cf2722af831cf00bcb5202381aa4","target":"record","created_at":"2026-05-17T23:39:04Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"280a24a132efba1de55e29405797d479ae87fbabcf2744fb8d296e8250bdacee","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-05T10:27:19Z","title_canon_sha256":"2e0bcb19e80d0349d0fdc24f12870751dcde889c95cf9f568bd49e86f84ffe28"},"schema_version":"1.0","source":{"id":"2601.01972","kind":"arxiv","version":4}},"canonical_sha256":"baf14f7782e46973f19c9707cf024bcedd6828d1905a8257862e7a9a3bd2fd70","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"baf14f7782e46973f19c9707cf024bcedd6828d1905a8257862e7a9a3bd2fd70","first_computed_at":"2026-05-17T23:39:04.524425Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:04.524425Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0tDBI8cBM6SXegtDfJhGBiKwcCR3aV8J6gMM/OVpWiBo4G4q5LBSQRTvj22YhToRkxIgPKYFa5WSCU8KAUnzAQ==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:04.525096Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.01972","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5025630261828b3b1181c5bfb1540da0fe92cf2722af831cf00bcb5202381aa4","sha256:8639f44d3bf9aa2ba5ee3f144404164d02bbb1574ad490ef3a5d7e94262e535c"],"state_sha256":"8d1ab40c4faa8cbb00922e3a138fc2e53b2cadb68135571b6553aa0149883e16"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Fi9v/xT3rnbpEpTBhicAMvYgGYNh5cS76Xoma6YYqhVHT+bVUvZNaKExMmevp0y4bAe1oq6045ZPuM7V7JcpAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-02T10:24:02.178951Z","bundle_sha256":"a2308bf31386f21fd58d9542f69d94834916d9a5cee8f20aa9fcbc0ed8eae632"}}