{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ECALI4OFANEXYE2OSNPJB2AP4N","short_pith_number":"pith:ECALI4OF","schema_version":"1.0","canonical_sha256":"2080b471c503497c134e935e90e80fe3507589c75eabd5d1635af5fc842f3213","source":{"kind":"arxiv","id":"2605.14084","version":1},"attestation_state":"computed","paper":{"title":"CRANE: Constrained Reasoning Injection for Code Agents via Nullspace Editing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"CRANE injects selected reasoning directions from thinking checkpoints into instruct models by editing their parameter nullspace, raising code agent success rates on benchmarks while keeping tool-use efficiency.","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.SE","authors_text":"Michele Merler, Mingzhi Zhu, Raju Pavuluri, Stacy Patterson","submitted_at":"2026-05-13T20:09:35Z","abstract_excerpt":"Code agents must both reason over long-horizon repository state and obey strict tool-use protocols. In paired Instruct/Thinking checkpoints, these capabilities are complementary but misaligned. The Instruct model is concise and tool-disciplined, whereas the Thinking model offers stronger planning and recovery behavior but often over-deliberates and degrades agent performance. We present CRANE (Constrained Reasoning Injection for Code Agents via Nullspace Editing), a training-free parameter-editing method that treats the Thinking-Instruct delta as a directional pool of candidate reasoning edits"},"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":true,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.14084","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-13T20:09:35Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"6aba564456d26590cf6e7e2caa1c38f0c4de2add3d160cc377eeb28ed81ddd11","abstract_canon_sha256":"5f018de7b629e31317483c2263c951ff3f6f73fd51bd0d2334ccf054c10150e2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:12.281132Z","signature_b64":"vRytmZithn4A0Tv1df48mXvD23g7fGxb7pRKKx0UhgmmHie32rfqUTdLd9GLqfaDqz5P1qLI6p46fmcKBYa9CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2080b471c503497c134e935e90e80fe3507589c75eabd5d1635af5fc842f3213","last_reissued_at":"2026-05-17T23:39:12.280511Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:12.280511Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CRANE: Constrained Reasoning Injection for Code Agents via Nullspace Editing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"CRANE injects selected reasoning directions from thinking checkpoints into instruct models by editing their parameter nullspace, raising code agent success rates on benchmarks while keeping tool-use efficiency.","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.SE","authors_text":"Michele Merler, Mingzhi Zhu, Raju Pavuluri, Stacy Patterson","submitted_at":"2026-05-13T20:09:35Z","abstract_excerpt":"Code agents must both reason over long-horizon repository state and obey strict tool-use protocols. In paired Instruct/Thinking checkpoints, these capabilities are complementary but misaligned. The Instruct model is concise and tool-disciplined, whereas the Thinking model offers stronger planning and recovery behavior but often over-deliberates and degrades agent performance. We present CRANE (Constrained Reasoning Injection for Code Agents via Nullspace Editing), a training-free parameter-editing method that treats the Thinking-Instruct delta as a directional pool of candidate reasoning edits"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"By merging paired Instruct and Thinking checkpoints, CRANE delivers strong gains over either individual model while preserving Instruct-level efficiency: on Roo-Eval it achieves pass1 of 66.2% (+19.5%) for Qwen3-30B-A3B and 81.5% (+8.7%) for Qwen3-Next-80B-A3B; on SWE-bench-Verified it resolves up to 14 additional instances at both scales (122/500 and 180/500); and on Terminal-Bench v2 it improves pass1/pass5 by up to 2.3%/7.8%, reaching 7.6%/17.9% and 14.8%/30.3%, respectively, consistently outperforming alternative merging strategies across all three benchmarks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the Thinking-Instruct delta vector contains directional information that can be selectively transferred via magnitude thresholding, Conservative Taylor Gate, and Graduated Sigmoidal Projection without degrading tool-use protocols or introducing new failure modes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CRANE merges Instruct and Thinking model checkpoints via constrained nullspace editing to improve code agent reasoning and benchmark performance without retraining.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"CRANE injects selected reasoning directions from thinking checkpoints into instruct models by editing their parameter nullspace, raising code agent success rates on benchmarks while keeping tool-use efficiency.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"616447d40d0cd1ec1642b09cc23906aca6d39df32be06afc62f5732e36c6ad77"},"source":{"id":"2605.14084","kind":"arxiv","version":1},"verdict":{"id":"df310626-5be2-45b9-8a4e-e48893b0fea6","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T04:42:27.537662Z","strongest_claim":"By merging paired Instruct and Thinking checkpoints, CRANE delivers strong gains over either individual model while preserving Instruct-level efficiency: on Roo-Eval it achieves pass1 of 66.2% (+19.5%) for Qwen3-30B-A3B and 81.5% (+8.7%) for Qwen3-Next-80B-A3B; on SWE-bench-Verified it resolves up to 14 additional instances at both scales (122/500 and 180/500); and on Terminal-Bench v2 it improves pass1/pass5 by up to 2.3%/7.8%, reaching 7.6%/17.9% and 14.8%/30.3%, respectively, consistently outperforming alternative merging strategies across all three benchmarks.","one_line_summary":"CRANE merges Instruct and Thinking model checkpoints via constrained nullspace editing to improve code agent reasoning and benchmark performance without retraining.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the Thinking-Instruct delta vector contains directional information that can be selectively transferred via magnitude thresholding, Conservative Taylor Gate, and Graduated Sigmoidal Projection without degrading tool-use protocols or introducing new failure modes.","pith_extraction_headline":"CRANE injects selected reasoning directions from thinking checkpoints into instruct models by editing their parameter nullspace, raising code agent success rates on benchmarks while keeping tool-use efficiency."},"references":{"count":21,"sample":[{"doi":"","year":2026,"title":"Qwen3-Coder-Next Technical Report","work_id":"ad966e68-641d-4b33-a9da-57cf741f35a6","ref_index":1,"cited_arxiv_id":"2603.00729","is_internal_anchor":true},{"doi":"","year":null,"title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","work_id":"e6b75ad5-2877-4168-97c8-710407094d20","ref_index":2,"cited_arxiv_id":"2501.12948","is_internal_anchor":true},{"doi":"","year":null,"title":"Lorahub: Efficient cross-task generalization via dynamic lora composition","work_id":"66ee831a-2b33-423d-87ec-ba52b341a21b","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"SWE-bench: Can Language Models Resolve Real-World GitHub Issues?","work_id":"d0effe15-a689-441a-8e3f-ea35f1c4e4b1","ref_index":4,"cited_arxiv_id":"2310.06770","is_internal_anchor":true},{"doi":"","year":null,"title":"arXiv preprint arXiv:2505.22113 , year=","work_id":"1eb7cd22-ed7f-48ef-be86-59775a274675","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":21,"snapshot_sha256":"8e50bd1b8a5fe7bf2a0bf23fcc1a9212cdf947d8adc8cc87d7fdd9996534443c","internal_anchors":7},"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":"2605.14084","created_at":"2026-05-17T23:39:12.280599+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.14084v1","created_at":"2026-05-17T23:39:12.280599+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14084","created_at":"2026-05-17T23:39:12.280599+00:00"},{"alias_kind":"pith_short_12","alias_value":"ECALI4OFANEX","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"ECALI4OFANEXYE2O","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"ECALI4OF","created_at":"2026-05-18T12:33:37.589309+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/ECALI4OFANEXYE2OSNPJB2AP4N","json":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N.json","graph_json":"https://pith.science/api/pith-number/ECALI4OFANEXYE2OSNPJB2AP4N/graph.json","events_json":"https://pith.science/api/pith-number/ECALI4OFANEXYE2OSNPJB2AP4N/events.json","paper":"https://pith.science/paper/ECALI4OF"},"agent_actions":{"view_html":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N","download_json":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N.json","view_paper":"https://pith.science/paper/ECALI4OF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.14084&json=true","fetch_graph":"https://pith.science/api/pith-number/ECALI4OFANEXYE2OSNPJB2AP4N/graph.json","fetch_events":"https://pith.science/api/pith-number/ECALI4OFANEXYE2OSNPJB2AP4N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N/action/storage_attestation","attest_author":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N/action/author_attestation","sign_citation":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N/action/citation_signature","submit_replication":"https://pith.science/pith/ECALI4OFANEXYE2OSNPJB2AP4N/action/replication_record"}},"created_at":"2026-05-17T23:39:12.280599+00:00","updated_at":"2026-05-17T23:39:12.280599+00:00"}