{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:KP5PER6UAXLZ4GYTXB4626NNTR","short_pith_number":"pith:KP5PER6U","schema_version":"1.0","canonical_sha256":"53faf247d405d79e1b13b879ed79ad9c4dcd9197c0e37484a932ef7883ce9a06","source":{"kind":"arxiv","id":"2606.12364","version":1},"attestation_state":"computed","paper":{"title":"On Subquadratic Architectures: From Applications to Principles","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anamaria-Roberta Hartl, David Stap, G\\\"unter Klambauer, Levente Z\\'olyomi, Lukas Hauzenberger, Niklas Schmidinger, Pieter-Jan Hoedt, Sebastian B\\\"ock, Sepp Hochreiter","submitted_at":"2026-06-10T17:33:55Z","abstract_excerpt":"Transformers dominate modern sequence modeling, but their quadratic attention incurs substantial computational cost. Subquadratic architectures offer a scalable alternative. However, it remains unclear which designs yield the most effective sequence models. We compare three leading approaches: xLSTM, Mamba-2, and Gated DeltaNet. We evaluate these models on tasks with complex dependencies: (1) code-model pre-training, (2) distillation of code models from large language models, and (3) pre-training of time-series foundation models. Across these settings, xLSTM delivers the strongest overall perf"},"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":"2606.12364","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-10T17:33:55Z","cross_cats_sorted":[],"title_canon_sha256":"0763b756831c10a022226a7261282fa8d254ff071fcd1801442573d3a39277d6","abstract_canon_sha256":"86b96c51d581277fc241d497909277e86e4bcda8cb2debc57f6b7c1bfc0b143a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-11T02:09:48.519662Z","signature_b64":"RONO0vDyYFpKiHnbYtQI6ypcEDpk7Wqy2Mbg1BcB6qWSyFj5ljMfjX37uCQuoxjQ5bxAVhvUtHNfHYo9SbDuBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"53faf247d405d79e1b13b879ed79ad9c4dcd9197c0e37484a932ef7883ce9a06","last_reissued_at":"2026-06-11T02:09:48.518948Z","signature_status":"signed_v1","first_computed_at":"2026-06-11T02:09:48.518948Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On Subquadratic Architectures: From Applications to Principles","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anamaria-Roberta Hartl, David Stap, G\\\"unter Klambauer, Levente Z\\'olyomi, Lukas Hauzenberger, Niklas Schmidinger, Pieter-Jan Hoedt, Sebastian B\\\"ock, Sepp Hochreiter","submitted_at":"2026-06-10T17:33:55Z","abstract_excerpt":"Transformers dominate modern sequence modeling, but their quadratic attention incurs substantial computational cost. Subquadratic architectures offer a scalable alternative. However, it remains unclear which designs yield the most effective sequence models. We compare three leading approaches: xLSTM, Mamba-2, and Gated DeltaNet. We evaluate these models on tasks with complex dependencies: (1) code-model pre-training, (2) distillation of code models from large language models, and (3) pre-training of time-series foundation models. Across these settings, xLSTM delivers the strongest overall perf"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12364","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/2606.12364/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":"2606.12364","created_at":"2026-06-11T02:09:48.519043+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.12364v1","created_at":"2026-06-11T02:09:48.519043+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12364","created_at":"2026-06-11T02:09:48.519043+00:00"},{"alias_kind":"pith_short_12","alias_value":"KP5PER6UAXLZ","created_at":"2026-06-11T02:09:48.519043+00:00"},{"alias_kind":"pith_short_16","alias_value":"KP5PER6UAXLZ4GYT","created_at":"2026-06-11T02:09:48.519043+00:00"},{"alias_kind":"pith_short_8","alias_value":"KP5PER6U","created_at":"2026-06-11T02:09:48.519043+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR","json":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR.json","graph_json":"https://pith.science/api/pith-number/KP5PER6UAXLZ4GYTXB4626NNTR/graph.json","events_json":"https://pith.science/api/pith-number/KP5PER6UAXLZ4GYTXB4626NNTR/events.json","paper":"https://pith.science/paper/KP5PER6U"},"agent_actions":{"view_html":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR","download_json":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR.json","view_paper":"https://pith.science/paper/KP5PER6U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.12364&json=true","fetch_graph":"https://pith.science/api/pith-number/KP5PER6UAXLZ4GYTXB4626NNTR/graph.json","fetch_events":"https://pith.science/api/pith-number/KP5PER6UAXLZ4GYTXB4626NNTR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR/action/storage_attestation","attest_author":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR/action/author_attestation","sign_citation":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR/action/citation_signature","submit_replication":"https://pith.science/pith/KP5PER6UAXLZ4GYTXB4626NNTR/action/replication_record"}},"created_at":"2026-06-11T02:09:48.519043+00:00","updated_at":"2026-06-11T02:09:48.519043+00:00"}