{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RZV2D4YAJUZQXPOHOUI7QXZPMZ","short_pith_number":"pith:RZV2D4YA","schema_version":"1.0","canonical_sha256":"8e6ba1f3004d330bbdc77511f85f2f66565fc6bc21938354d8afecda7ccb6638","source":{"kind":"arxiv","id":"2504.15777","version":1},"attestation_state":"computed","paper":{"title":"Tina: Tiny Reasoning Models via LoRA","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Enes Burak Bilgin, Julian Asilis, Ollie Liu, \\\"Omer Faruk Akg\\\"ul, Shangshang Wang, Willie Neiswanger","submitted_at":"2025-04-22T10:38:00Z","abstract_excerpt":"How cost-effectively can strong reasoning abilities be achieved in language models? Driven by this fundamental question, we present Tina, a family of tiny reasoning models achieved with high cost-efficiency. Notably, Tina demonstrates that substantial reasoning performance can be developed using only minimal resources, by applying parameter-efficient updates during reinforcement learning (RL), using low-rank adaptation (LoRA), to an already tiny 1.5B parameter base model. This minimalist approach produces models that achieve reasoning performance which is competitive with, and sometimes surpas"},"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":"2504.15777","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-22T10:38:00Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"20423242d72c7fbab69be6d0b3c5d7b8e1a00242aa36a97997357d5bcef21c9a","abstract_canon_sha256":"99109c07248ba3b8e22cb6b1c34bb127e8097884aad705d6bd4f0e232d6fd882"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:20:11.566597Z","signature_b64":"v5k+6P/rheZtLw7VirvorFLfck5xov4FId8zGXxvfBa1+p/fVOB1N7ZltWp6Wykij3X+FvEJxgDq/CyyAJofDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8e6ba1f3004d330bbdc77511f85f2f66565fc6bc21938354d8afecda7ccb6638","last_reissued_at":"2026-07-05T11:20:11.566012Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:20:11.566012Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Tina: Tiny Reasoning Models via LoRA","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Enes Burak Bilgin, Julian Asilis, Ollie Liu, \\\"Omer Faruk Akg\\\"ul, Shangshang Wang, Willie Neiswanger","submitted_at":"2025-04-22T10:38:00Z","abstract_excerpt":"How cost-effectively can strong reasoning abilities be achieved in language models? Driven by this fundamental question, we present Tina, a family of tiny reasoning models achieved with high cost-efficiency. Notably, Tina demonstrates that substantial reasoning performance can be developed using only minimal resources, by applying parameter-efficient updates during reinforcement learning (RL), using low-rank adaptation (LoRA), to an already tiny 1.5B parameter base model. This minimalist approach produces models that achieve reasoning performance which is competitive with, and sometimes surpas"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.15777","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/2504.15777/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":"2504.15777","created_at":"2026-07-05T11:20:11.566072+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.15777v1","created_at":"2026-07-05T11:20:11.566072+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.15777","created_at":"2026-07-05T11:20:11.566072+00:00"},{"alias_kind":"pith_short_12","alias_value":"RZV2D4YAJUZQ","created_at":"2026-07-05T11:20:11.566072+00:00"},{"alias_kind":"pith_short_16","alias_value":"RZV2D4YAJUZQXPOH","created_at":"2026-07-05T11:20:11.566072+00:00"},{"alias_kind":"pith_short_8","alias_value":"RZV2D4YA","created_at":"2026-07-05T11:20:11.566072+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.12883","citing_title":"The Hidden Power of Scaling Factor in LoRA Optimization","ref_index":123,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06241","citing_title":"Rethinking RL for LLM Reasoning: It's Sparse Policy Selection, Not Capability Learning","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06241","citing_title":"Rethinking RL for LLM Reasoning: It's Sparse Policy Selection, Not Capability Learning","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21905","citing_title":"Low-Rank Adaptation Redux for Large Models","ref_index":199,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18936","citing_title":"Fine-Tuning Small Reasoning Models for Quantum Field Theory","ref_index":217,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12160","citing_title":"PubSwap: Public-Data Off-Policy Coordination for Federated RLVR","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05426","citing_title":"ALTO: Adaptive LoRA Tuning and Orchestration for Heterogeneous LoRA Training Workloads","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18381","citing_title":"Learning from Less: Measuring the Effectiveness of RLVR in Low Data and Compute Regimes","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ","json":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ.json","graph_json":"https://pith.science/api/pith-number/RZV2D4YAJUZQXPOHOUI7QXZPMZ/graph.json","events_json":"https://pith.science/api/pith-number/RZV2D4YAJUZQXPOHOUI7QXZPMZ/events.json","paper":"https://pith.science/paper/RZV2D4YA"},"agent_actions":{"view_html":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ","download_json":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ.json","view_paper":"https://pith.science/paper/RZV2D4YA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.15777&json=true","fetch_graph":"https://pith.science/api/pith-number/RZV2D4YAJUZQXPOHOUI7QXZPMZ/graph.json","fetch_events":"https://pith.science/api/pith-number/RZV2D4YAJUZQXPOHOUI7QXZPMZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ/action/storage_attestation","attest_author":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ/action/author_attestation","sign_citation":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ/action/citation_signature","submit_replication":"https://pith.science/pith/RZV2D4YAJUZQXPOHOUI7QXZPMZ/action/replication_record"}},"created_at":"2026-07-05T11:20:11.566072+00:00","updated_at":"2026-07-05T11:20:11.566072+00:00"}