{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6SHE7WMIJ3A3BIDZJT6WYVAPAL","short_pith_number":"pith:6SHE7WMI","schema_version":"1.0","canonical_sha256":"f48e4fd9884ec1b0a0794cfd6c540f02d0af360c63c5fd40e2b9269961deecd0","source":{"kind":"arxiv","id":"2410.11135","version":1},"attestation_state":"computed","paper":{"title":"Mimetic Initialization Helps State Space Models Learn to Recall","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Asher Trockman, Hrayr Harutyunyan, J. Zico Kolter, Sanjiv Kumar, Srinadh Bhojanapalli","submitted_at":"2024-10-14T23:17:46Z","abstract_excerpt":"Recent work has shown that state space models such as Mamba are significantly worse than Transformers on recall-based tasks due to the fact that their state size is constant with respect to their input sequence length. But in practice, state space models have fairly large state sizes, and we conjecture that they should be able to perform much better at these tasks than previously reported. We investigate whether their poor copying and recall performance could be due in part to training difficulties rather than fundamental capacity constraints. Based on observations of their \"attention\" maps, w"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2410.11135","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-14T23:17:46Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"34c87da9899f7f3af1d9afc8e25da2142f010f5538feb735b9749115e7b7af76","abstract_canon_sha256":"a403d84824960d93dcf3288dc6a471b3b4b0c62ce7192b8690c3edb7bd55fb1c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:20:29.745454Z","signature_b64":"qK0Dlz7q9KKrhy7VphG57jGISOwG/s/AJC22xZSigjg5CNF8ZQbjgDgDnbvb8UMxqhang38nS5E2xUOG9nK0Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f48e4fd9884ec1b0a0794cfd6c540f02d0af360c63c5fd40e2b9269961deecd0","last_reissued_at":"2026-07-05T09:20:29.744957Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:20:29.744957Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mimetic Initialization Helps State Space Models Learn to Recall","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Asher Trockman, Hrayr Harutyunyan, J. Zico Kolter, Sanjiv Kumar, Srinadh Bhojanapalli","submitted_at":"2024-10-14T23:17:46Z","abstract_excerpt":"Recent work has shown that state space models such as Mamba are significantly worse than Transformers on recall-based tasks due to the fact that their state size is constant with respect to their input sequence length. But in practice, state space models have fairly large state sizes, and we conjecture that they should be able to perform much better at these tasks than previously reported. We investigate whether their poor copying and recall performance could be due in part to training difficulties rather than fundamental capacity constraints. Based on observations of their \"attention\" maps, w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.11135","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2410.11135/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2410.11135","created_at":"2026-07-05T09:20:29.745022+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.11135v1","created_at":"2026-07-05T09:20:29.745022+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.11135","created_at":"2026-07-05T09:20:29.745022+00:00"},{"alias_kind":"pith_short_12","alias_value":"6SHE7WMIJ3A3","created_at":"2026-07-05T09:20:29.745022+00:00"},{"alias_kind":"pith_short_16","alias_value":"6SHE7WMIJ3A3BIDZ","created_at":"2026-07-05T09:20:29.745022+00:00"},{"alias_kind":"pith_short_8","alias_value":"6SHE7WMI","created_at":"2026-07-05T09:20:29.745022+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.21070","citing_title":"Towards Understanding Self-Pretraining for Sequence Classification","ref_index":121,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL","json":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL.json","graph_json":"https://pith.science/api/pith-number/6SHE7WMIJ3A3BIDZJT6WYVAPAL/graph.json","events_json":"https://pith.science/api/pith-number/6SHE7WMIJ3A3BIDZJT6WYVAPAL/events.json","paper":"https://pith.science/paper/6SHE7WMI"},"agent_actions":{"view_html":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL","download_json":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL.json","view_paper":"https://pith.science/paper/6SHE7WMI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.11135&json=true","fetch_graph":"https://pith.science/api/pith-number/6SHE7WMIJ3A3BIDZJT6WYVAPAL/graph.json","fetch_events":"https://pith.science/api/pith-number/6SHE7WMIJ3A3BIDZJT6WYVAPAL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL/action/storage_attestation","attest_author":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL/action/author_attestation","sign_citation":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL/action/citation_signature","submit_replication":"https://pith.science/pith/6SHE7WMIJ3A3BIDZJT6WYVAPAL/action/replication_record"}},"created_at":"2026-07-05T09:20:29.745022+00:00","updated_at":"2026-07-05T09:20:29.745022+00:00"}