{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:AJXGYJFPV7EPCXUG7U2OKXG2W2","short_pith_number":"pith:AJXGYJFP","schema_version":"1.0","canonical_sha256":"026e6c24afafc8f15e86fd34e55cdab6aeae333a7afb012d2ff43ec4522b2a9c","source":{"kind":"arxiv","id":"2606.09059","version":1},"attestation_state":"computed","paper":{"title":"Stage-1 Controls the Entropy Regime, Not the Outcome","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Jianxiong Shen","submitted_at":"2026-06-08T05:49:09Z","abstract_excerpt":"Two-stage post-training -- a Stage-1 warm-start (supervised fine-tuning, SFT, or on-policy distillation, OPD) followed by Stage-2 reinforcement learning (RL) -- is increasingly used for vision-language models (VLMs). We ask what Stage-1 actually controls in a small-data study using Qwen2.5-VL-7B with a same-modality 72B VLM teacher for OPD. First, the three warm-starts reach a narrow $53$--$54\\%$ band on Geometry3K internal validation, consistent with the narrow range reported by recent specialized methods; this setup provides little evidence that Stage-1 changes the in-domain endpoint. Second"},"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":"2606.09059","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T05:49:09Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"e017da39fff2635e40f890dab023a2bad847bda90e94f5da6cee10cdf69bf103","abstract_canon_sha256":"9cdb18e1eb43b73583ff968b4462fdf97057a36763e8c2e95bc79884ac6b9d55"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:07:56.233257Z","signature_b64":"6i9qlvbuvy39Flq6gv+nInU+mAI2yEzJQ3GzY3YEChJmOjfm8Qi9J2aVLp67rqTLJ0ny5Tf2kze4DbfzVaEfBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"026e6c24afafc8f15e86fd34e55cdab6aeae333a7afb012d2ff43ec4522b2a9c","last_reissued_at":"2026-06-09T02:07:56.232465Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:07:56.232465Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Stage-1 Controls the Entropy Regime, Not the Outcome","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Jianxiong Shen","submitted_at":"2026-06-08T05:49:09Z","abstract_excerpt":"Two-stage post-training -- a Stage-1 warm-start (supervised fine-tuning, SFT, or on-policy distillation, OPD) followed by Stage-2 reinforcement learning (RL) -- is increasingly used for vision-language models (VLMs). We ask what Stage-1 actually controls in a small-data study using Qwen2.5-VL-7B with a same-modality 72B VLM teacher for OPD. First, the three warm-starts reach a narrow $53$--$54\\%$ band on Geometry3K internal validation, consistent with the narrow range reported by recent specialized methods; this setup provides little evidence that Stage-1 changes the in-domain endpoint. Second"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09059","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/2606.09059/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":"2606.09059","created_at":"2026-06-09T02:07:56.232592+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.09059v1","created_at":"2026-06-09T02:07:56.232592+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09059","created_at":"2026-06-09T02:07:56.232592+00:00"},{"alias_kind":"pith_short_12","alias_value":"AJXGYJFPV7EP","created_at":"2026-06-09T02:07:56.232592+00:00"},{"alias_kind":"pith_short_16","alias_value":"AJXGYJFPV7EPCXUG","created_at":"2026-06-09T02:07:56.232592+00:00"},{"alias_kind":"pith_short_8","alias_value":"AJXGYJFP","created_at":"2026-06-09T02:07:56.232592+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/AJXGYJFPV7EPCXUG7U2OKXG2W2","json":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2.json","graph_json":"https://pith.science/api/pith-number/AJXGYJFPV7EPCXUG7U2OKXG2W2/graph.json","events_json":"https://pith.science/api/pith-number/AJXGYJFPV7EPCXUG7U2OKXG2W2/events.json","paper":"https://pith.science/paper/AJXGYJFP"},"agent_actions":{"view_html":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2","download_json":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2.json","view_paper":"https://pith.science/paper/AJXGYJFP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.09059&json=true","fetch_graph":"https://pith.science/api/pith-number/AJXGYJFPV7EPCXUG7U2OKXG2W2/graph.json","fetch_events":"https://pith.science/api/pith-number/AJXGYJFPV7EPCXUG7U2OKXG2W2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2/action/storage_attestation","attest_author":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2/action/author_attestation","sign_citation":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2/action/citation_signature","submit_replication":"https://pith.science/pith/AJXGYJFPV7EPCXUG7U2OKXG2W2/action/replication_record"}},"created_at":"2026-06-09T02:07:56.232592+00:00","updated_at":"2026-06-09T02:07:56.232592+00:00"}