{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BBSHL5QNNERYQQD245AC6GQFAF","short_pith_number":"pith:BBSHL5QN","schema_version":"1.0","canonical_sha256":"086475f60d692388407ae7402f1a050179e2ade0ae9daf5fdd42f82cc2e75ac0","source":{"kind":"arxiv","id":"2507.20984","version":2},"attestation_state":"computed","paper":{"title":"SmallThinker: A Family of Efficient Large Language Models Natively Trained for Local Deployment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Bo Wen, Chengrong Tian, Dongliang Wei, Feiyang Chen, Guangshuo Qin, Haibo Chen, Hangyu Liang, Jianxiang Gao, Junchen Liu, Longyu Zhao, Xinrui Zheng, Yixin Song, Zeyu Mi, Zhenliang Xue","submitted_at":"2025-07-28T16:45:14Z","abstract_excerpt":"While frontier large language models (LLMs) continue to push capability boundaries, their deployment remains confined to GPU-powered cloud infrastructure. We challenge this paradigm with SmallThinker, a family of LLMs natively designed - not adapted - for the unique constraints of local devices: weak computational power, limited memory, and slow storage. Unlike traditional approaches that mainly compress existing models built for clouds, we architect SmallThinker from the ground up to thrive within these limitations. Our innovation lies in a deployment-aware architecture that transforms constr"},"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":"2507.20984","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-28T16:45:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ee2d25112cb5d068f11e55b26376d05c461fbe9826fb894c946e6f935a10c9a6","abstract_canon_sha256":"99c6cd278388a414f93707e0af20e8033f34daa6f6fd9e623edd81791e715c3c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:45:38.621933Z","signature_b64":"cqfkHQVVqAsWxinat2Edp/PPFTnqozEKUXHL8M/iTVAYh6qWjEm+fnkEatjSMkoCKcykXMhnOMgHiji31zYfDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"086475f60d692388407ae7402f1a050179e2ade0ae9daf5fdd42f82cc2e75ac0","last_reissued_at":"2026-07-05T11:45:38.621444Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:45:38.621444Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SmallThinker: A Family of Efficient Large Language Models Natively Trained for Local Deployment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Bo Wen, Chengrong Tian, Dongliang Wei, Feiyang Chen, Guangshuo Qin, Haibo Chen, Hangyu Liang, Jianxiang Gao, Junchen Liu, Longyu Zhao, Xinrui Zheng, Yixin Song, Zeyu Mi, Zhenliang Xue","submitted_at":"2025-07-28T16:45:14Z","abstract_excerpt":"While frontier large language models (LLMs) continue to push capability boundaries, their deployment remains confined to GPU-powered cloud infrastructure. We challenge this paradigm with SmallThinker, a family of LLMs natively designed - not adapted - for the unique constraints of local devices: weak computational power, limited memory, and slow storage. Unlike traditional approaches that mainly compress existing models built for clouds, we architect SmallThinker from the ground up to thrive within these limitations. Our innovation lies in a deployment-aware architecture that transforms constr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.20984","kind":"arxiv","version":2},"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/2507.20984/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":"2507.20984","created_at":"2026-07-05T11:45:38.621503+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.20984v2","created_at":"2026-07-05T11:45:38.621503+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.20984","created_at":"2026-07-05T11:45:38.621503+00:00"},{"alias_kind":"pith_short_12","alias_value":"BBSHL5QNNERY","created_at":"2026-07-05T11:45:38.621503+00:00"},{"alias_kind":"pith_short_16","alias_value":"BBSHL5QNNERYQQD2","created_at":"2026-07-05T11:45:38.621503+00:00"},{"alias_kind":"pith_short_8","alias_value":"BBSHL5QN","created_at":"2026-07-05T11:45:38.621503+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.27281","citing_title":"Resource-Aware Neuro-Symbolic Reasoning for Local Small Language Models","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF","json":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF.json","graph_json":"https://pith.science/api/pith-number/BBSHL5QNNERYQQD245AC6GQFAF/graph.json","events_json":"https://pith.science/api/pith-number/BBSHL5QNNERYQQD245AC6GQFAF/events.json","paper":"https://pith.science/paper/BBSHL5QN"},"agent_actions":{"view_html":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF","download_json":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF.json","view_paper":"https://pith.science/paper/BBSHL5QN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.20984&json=true","fetch_graph":"https://pith.science/api/pith-number/BBSHL5QNNERYQQD245AC6GQFAF/graph.json","fetch_events":"https://pith.science/api/pith-number/BBSHL5QNNERYQQD245AC6GQFAF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF/action/storage_attestation","attest_author":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF/action/author_attestation","sign_citation":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF/action/citation_signature","submit_replication":"https://pith.science/pith/BBSHL5QNNERYQQD245AC6GQFAF/action/replication_record"}},"created_at":"2026-07-05T11:45:38.621503+00:00","updated_at":"2026-07-05T11:45:38.621503+00:00"}