{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:DHL6PRYNZ7HHIMTDSYEOGK36U3","short_pith_number":"pith:DHL6PRYN","schema_version":"1.0","canonical_sha256":"19d7e7c70dcfce7432639608e32b7ea6f7e2e6fe998a9f3096cbc3f416cf5ed3","source":{"kind":"arxiv","id":"2601.20789","version":3},"attestation_state":"computed","paper":{"title":"SERA: Soft-Verified Efficient Repository Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","cs.SE"],"primary_cat":"cs.CL","authors_text":"Ali Farhadi, Daniel Tormoen, Ethan Shen, Saurabh Shah, Tim Dettmers","submitted_at":"2026-01-28T17:27:08Z","abstract_excerpt":"Open-weight coding agents should hold a fundamental advantage over closed-source systems because they can specialize to private codebases, encoding repository-specific information directly in their weights. Yet the cost and complexity of training has kept this advantage theoretical until now. We present Soft-Verified Efficient Repository Agents (SERA), an efficient method for training coding agents that enables the rapid and cheap creation of agents specialized to private codebases. Using Soft Verified Generation (SVG), we generate thousands of trajectories from any code repository, without re"},"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":"2601.20789","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-28T17:27:08Z","cross_cats_sorted":["cs.LG","cs.SE"],"title_canon_sha256":"e48d16b1a34b6da23ba5500df7f7bab64e1308fbb32c65cc4dd1cd59d75349e6","abstract_canon_sha256":"6ff6a88a4008e2eabd5b3ae6b3a805b5cf15545b195f705f127833fae157f9aa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:02:30.651920Z","signature_b64":"fAfNcetAG8U3XAsviKLRn8Eby+NC6+BK5U5SHeeldvFDwKleOcNnv1kTioTEW89K/+nwMi02No72PtWE/ldKCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"19d7e7c70dcfce7432639608e32b7ea6f7e2e6fe998a9f3096cbc3f416cf5ed3","last_reissued_at":"2026-06-01T01:02:30.650991Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:02:30.650991Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SERA: Soft-Verified Efficient Repository Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","cs.SE"],"primary_cat":"cs.CL","authors_text":"Ali Farhadi, Daniel Tormoen, Ethan Shen, Saurabh Shah, Tim Dettmers","submitted_at":"2026-01-28T17:27:08Z","abstract_excerpt":"Open-weight coding agents should hold a fundamental advantage over closed-source systems because they can specialize to private codebases, encoding repository-specific information directly in their weights. Yet the cost and complexity of training has kept this advantage theoretical until now. We present Soft-Verified Efficient Repository Agents (SERA), an efficient method for training coding agents that enables the rapid and cheap creation of agents specialized to private codebases. Using Soft Verified Generation (SVG), we generate thousands of trajectories from any code repository, without re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.20789","kind":"arxiv","version":3},"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/2601.20789/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":"2601.20789","created_at":"2026-06-01T01:02:30.651113+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.20789v3","created_at":"2026-06-01T01:02:30.651113+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.20789","created_at":"2026-06-01T01:02:30.651113+00:00"},{"alias_kind":"pith_short_12","alias_value":"DHL6PRYNZ7HH","created_at":"2026-06-01T01:02:30.651113+00:00"},{"alias_kind":"pith_short_16","alias_value":"DHL6PRYNZ7HHIMTD","created_at":"2026-06-01T01:02:30.651113+00:00"},{"alias_kind":"pith_short_8","alias_value":"DHL6PRYN","created_at":"2026-06-01T01:02:30.651113+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2604.01496","citing_title":"From SWE-ZERO to SWE-HERO: Execution-free to Execution-based Fine-tuning for Software Engineering Agents","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09879","citing_title":"M2A: Synergizing Mathematical and Agentic Reasoning in Large Language Models","ref_index":29,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3","json":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3.json","graph_json":"https://pith.science/api/pith-number/DHL6PRYNZ7HHIMTDSYEOGK36U3/graph.json","events_json":"https://pith.science/api/pith-number/DHL6PRYNZ7HHIMTDSYEOGK36U3/events.json","paper":"https://pith.science/paper/DHL6PRYN"},"agent_actions":{"view_html":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3","download_json":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3.json","view_paper":"https://pith.science/paper/DHL6PRYN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.20789&json=true","fetch_graph":"https://pith.science/api/pith-number/DHL6PRYNZ7HHIMTDSYEOGK36U3/graph.json","fetch_events":"https://pith.science/api/pith-number/DHL6PRYNZ7HHIMTDSYEOGK36U3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3/action/storage_attestation","attest_author":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3/action/author_attestation","sign_citation":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3/action/citation_signature","submit_replication":"https://pith.science/pith/DHL6PRYNZ7HHIMTDSYEOGK36U3/action/replication_record"}},"created_at":"2026-06-01T01:02:30.651113+00:00","updated_at":"2026-06-01T01:02:30.651113+00:00"}