{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CFOUDLSFABZLRJXZ46CLKYFONC","short_pith_number":"pith:CFOUDLSF","schema_version":"1.0","canonical_sha256":"115d41ae450072b8a6f9e784b560ae688c11c6671d48e6570caa6957979f2675","source":{"kind":"arxiv","id":"2505.22425","version":1},"attestation_state":"computed","paper":{"title":"Scaling Reasoning without Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Lingpeng Kong, Wei Wu, Xueliang Zhao","submitted_at":"2025-05-28T14:52:15Z","abstract_excerpt":"Large language models (LLMs) have made significant advances in complex reasoning tasks, yet they remain bottlenecked by two core challenges: architectural inefficiency due to reliance on Transformers, and a lack of structured fine-tuning for high-difficulty domains. We introduce \\ourmodel, an attention-free language model that addresses both issues through architectural and data-centric innovations. Built on the state space dual (SSD) layers of Mamba-2, our model eliminates the need for self-attention and key-value caching, enabling fixed-memory, constant-time inference. To train it for comple"},"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":"2505.22425","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-28T14:52:15Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"73eba11aba63998c1d24aee39d16d8567c4f1daa485ab47692ed92ef26c28d6a","abstract_canon_sha256":"96a695bf8f49269040c1ec286f1b6fd8766c8a267c626bc5e65e86e8f072d861"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:23.476565Z","signature_b64":"43J9y0a+kdr3j3X7o5OveMLLONvpA13usNIdBuxQV1c1Auz7FqGB1xXQvcgxUWDfrpJaxXLSuX/sRyBQkk/hDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"115d41ae450072b8a6f9e784b560ae688c11c6671d48e6570caa6957979f2675","last_reissued_at":"2026-07-05T11:11:23.476068Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:23.476068Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scaling Reasoning without Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Lingpeng Kong, Wei Wu, Xueliang Zhao","submitted_at":"2025-05-28T14:52:15Z","abstract_excerpt":"Large language models (LLMs) have made significant advances in complex reasoning tasks, yet they remain bottlenecked by two core challenges: architectural inefficiency due to reliance on Transformers, and a lack of structured fine-tuning for high-difficulty domains. We introduce \\ourmodel, an attention-free language model that addresses both issues through architectural and data-centric innovations. Built on the state space dual (SSD) layers of Mamba-2, our model eliminates the need for self-attention and key-value caching, enabling fixed-memory, constant-time inference. To train it for comple"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.22425","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/2505.22425/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":"2505.22425","created_at":"2026-07-05T11:11:23.476127+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.22425v1","created_at":"2026-07-05T11:11:23.476127+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.22425","created_at":"2026-07-05T11:11:23.476127+00:00"},{"alias_kind":"pith_short_12","alias_value":"CFOUDLSFABZL","created_at":"2026-07-05T11:11:23.476127+00:00"},{"alias_kind":"pith_short_16","alias_value":"CFOUDLSFABZLRJXZ","created_at":"2026-07-05T11:11:23.476127+00:00"},{"alias_kind":"pith_short_8","alias_value":"CFOUDLSF","created_at":"2026-07-05T11:11:23.476127+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.11634","citing_title":"Architecture-Aware Reinforcement Learning Makes Sliding-Window Attention Competitive in Math Reasoning","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2507.11687","citing_title":"MetaLint: Easy-to-Hard Generalization for Code Linting","ref_index":54,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC","json":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC.json","graph_json":"https://pith.science/api/pith-number/CFOUDLSFABZLRJXZ46CLKYFONC/graph.json","events_json":"https://pith.science/api/pith-number/CFOUDLSFABZLRJXZ46CLKYFONC/events.json","paper":"https://pith.science/paper/CFOUDLSF"},"agent_actions":{"view_html":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC","download_json":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC.json","view_paper":"https://pith.science/paper/CFOUDLSF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.22425&json=true","fetch_graph":"https://pith.science/api/pith-number/CFOUDLSFABZLRJXZ46CLKYFONC/graph.json","fetch_events":"https://pith.science/api/pith-number/CFOUDLSFABZLRJXZ46CLKYFONC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC/action/storage_attestation","attest_author":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC/action/author_attestation","sign_citation":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC/action/citation_signature","submit_replication":"https://pith.science/pith/CFOUDLSFABZLRJXZ46CLKYFONC/action/replication_record"}},"created_at":"2026-07-05T11:11:23.476127+00:00","updated_at":"2026-07-05T11:11:23.476127+00:00"}