{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XJQJ2WZNJSLQW62IRSHJTZGCDW","short_pith_number":"pith:XJQJ2WZN","schema_version":"1.0","canonical_sha256":"ba609d5b2d4c970b7b488c8e99e4c21daf6f0f387ce65d5ae3bbd932cde05837","source":{"kind":"arxiv","id":"2501.16394","version":1},"attestation_state":"computed","paper":{"title":"Transformer^-1: Input-Adaptive Computation for Resource-Constrained Deployment","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Fucheng Zhong, Jisen Jia, Lumen AI, Shihao Ji, Tengzhou No. 1 Middle School, Xu Tianhao, Zhaobo Wu, Zheyi Cao, Zihui Song","submitted_at":"2025-01-26T15:31:45Z","abstract_excerpt":"Addressing the resource waste caused by fixed computation paradigms in deep learning models under dynamic scenarios, this paper proposes a Transformer$^{-1}$ architecture based on the principle of deep adaptivity. This architecture achieves dynamic matching between input features and computational resources by establishing a joint optimization model for complexity and computation. Our core contributions include: (1) designing a two-layer control mechanism, composed of a complexity predictor and a reinforcement learning policy network, enabling end-to-end optimization of computation paths; (2) "},"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":"2501.16394","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-26T15:31:45Z","cross_cats_sorted":[],"title_canon_sha256":"440d695f70f8f98c43d1ee4f1e86a8fd9a22ab629475cbc18e091fb1fe5a8f17","abstract_canon_sha256":"06261a1d5ba3daa0d97cb46e1d221749bf5d9fe6d2ed42575620968495298f0f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:06:12.310243Z","signature_b64":"+Ceh/CTjOdO8R6RYlR8jA9zo9ZqW6jANEq1S7aBzlzNXwoLXBwvhvnocmUDWEkZaK6xSo/1XflzcPuMmduQzDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba609d5b2d4c970b7b488c8e99e4c21daf6f0f387ce65d5ae3bbd932cde05837","last_reissued_at":"2026-07-05T10:06:12.309767Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:06:12.309767Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Transformer^-1: Input-Adaptive Computation for Resource-Constrained Deployment","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Fucheng Zhong, Jisen Jia, Lumen AI, Shihao Ji, Tengzhou No. 1 Middle School, Xu Tianhao, Zhaobo Wu, Zheyi Cao, Zihui Song","submitted_at":"2025-01-26T15:31:45Z","abstract_excerpt":"Addressing the resource waste caused by fixed computation paradigms in deep learning models under dynamic scenarios, this paper proposes a Transformer$^{-1}$ architecture based on the principle of deep adaptivity. This architecture achieves dynamic matching between input features and computational resources by establishing a joint optimization model for complexity and computation. Our core contributions include: (1) designing a two-layer control mechanism, composed of a complexity predictor and a reinforcement learning policy network, enabling end-to-end optimization of computation paths; (2) "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.16394","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/2501.16394/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":"2501.16394","created_at":"2026-07-05T10:06:12.309825+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.16394v1","created_at":"2026-07-05T10:06:12.309825+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.16394","created_at":"2026-07-05T10:06:12.309825+00:00"},{"alias_kind":"pith_short_12","alias_value":"XJQJ2WZNJSLQ","created_at":"2026-07-05T10:06:12.309825+00:00"},{"alias_kind":"pith_short_16","alias_value":"XJQJ2WZNJSLQW62I","created_at":"2026-07-05T10:06:12.309825+00:00"},{"alias_kind":"pith_short_8","alias_value":"XJQJ2WZN","created_at":"2026-07-05T10:06:12.309825+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/XJQJ2WZNJSLQW62IRSHJTZGCDW","json":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW.json","graph_json":"https://pith.science/api/pith-number/XJQJ2WZNJSLQW62IRSHJTZGCDW/graph.json","events_json":"https://pith.science/api/pith-number/XJQJ2WZNJSLQW62IRSHJTZGCDW/events.json","paper":"https://pith.science/paper/XJQJ2WZN"},"agent_actions":{"view_html":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW","download_json":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW.json","view_paper":"https://pith.science/paper/XJQJ2WZN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.16394&json=true","fetch_graph":"https://pith.science/api/pith-number/XJQJ2WZNJSLQW62IRSHJTZGCDW/graph.json","fetch_events":"https://pith.science/api/pith-number/XJQJ2WZNJSLQW62IRSHJTZGCDW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW/action/storage_attestation","attest_author":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW/action/author_attestation","sign_citation":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW/action/citation_signature","submit_replication":"https://pith.science/pith/XJQJ2WZNJSLQW62IRSHJTZGCDW/action/replication_record"}},"created_at":"2026-07-05T10:06:12.309825+00:00","updated_at":"2026-07-05T10:06:12.309825+00:00"}