{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:J25YCLGLPWDZ7L7ZQGYV24DXK7","short_pith_number":"pith:J25YCLGL","schema_version":"1.0","canonical_sha256":"4ebb812ccb7d879faff981b15d707757c1b3fb9998f45a7d886a5e431c640373","source":{"kind":"arxiv","id":"2605.20246","version":1},"attestation_state":"computed","paper":{"title":"GROW: Aligning GRPO with State-Action Modeling for Open-World VLM Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jie Yang, Lechao Zhang, Shanzhe Lei, Wei Liu, Xin Tan, Xiongbin Wu, Xuhong Wang, Yuanjie Zheng, Zhihao Luo, Zhonglong Zheng","submitted_at":"2026-05-18T04:56:59Z","abstract_excerpt":"Recently, vision-language model (VLM) agents have shown promising progress in open-world tasks, where successful task completion often requires multiple turns of visual perception and action execution. However, existing methods still rely primarily on Supervised Fine-Tuning (SFT) with expert demonstrations, while the advanced reinforcement learning (RL) algorithm, specifically Group Relative Policy Optimization (GRPO), has not been effectively employed for multi-turn RL in these tasks because standard GRPO requires full trajectories as training samples which leads to excessively long context a"},"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":"2605.20246","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T04:56:59Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"259dc522ff9733faddfc41ff31c65c3858ca619b6137a572b02cffca6efe8c0f","abstract_canon_sha256":"ace2547a1c8dfdfd4f7ec25f0db54e1e80bb3444486960f0dbcf555d83126bfd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T00:04:22.569668Z","signature_b64":"qpe6y88qXb0ldUmF37pgQiFpR0zQ75gLwzbzq1MOndeMUZmww6X9W0ZU3snkXYEHB+A12XzWWPyL7/EX0jioBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4ebb812ccb7d879faff981b15d707757c1b3fb9998f45a7d886a5e431c640373","last_reissued_at":"2026-05-21T00:04:22.569171Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T00:04:22.569171Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GROW: Aligning GRPO with State-Action Modeling for Open-World VLM Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Jie Yang, Lechao Zhang, Shanzhe Lei, Wei Liu, Xin Tan, Xiongbin Wu, Xuhong Wang, Yuanjie Zheng, Zhihao Luo, Zhonglong Zheng","submitted_at":"2026-05-18T04:56:59Z","abstract_excerpt":"Recently, vision-language model (VLM) agents have shown promising progress in open-world tasks, where successful task completion often requires multiple turns of visual perception and action execution. However, existing methods still rely primarily on Supervised Fine-Tuning (SFT) with expert demonstrations, while the advanced reinforcement learning (RL) algorithm, specifically Group Relative Policy Optimization (GRPO), has not been effectively employed for multi-turn RL in these tasks because standard GRPO requires full trajectories as training samples which leads to excessively long context a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20246","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/2605.20246/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":"2605.20246","created_at":"2026-05-21T00:04:22.569236+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.20246v1","created_at":"2026-05-21T00:04:22.569236+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20246","created_at":"2026-05-21T00:04:22.569236+00:00"},{"alias_kind":"pith_short_12","alias_value":"J25YCLGLPWDZ","created_at":"2026-05-21T00:04:22.569236+00:00"},{"alias_kind":"pith_short_16","alias_value":"J25YCLGLPWDZ7L7Z","created_at":"2026-05-21T00:04:22.569236+00:00"},{"alias_kind":"pith_short_8","alias_value":"J25YCLGL","created_at":"2026-05-21T00:04:22.569236+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/J25YCLGLPWDZ7L7ZQGYV24DXK7","json":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7.json","graph_json":"https://pith.science/api/pith-number/J25YCLGLPWDZ7L7ZQGYV24DXK7/graph.json","events_json":"https://pith.science/api/pith-number/J25YCLGLPWDZ7L7ZQGYV24DXK7/events.json","paper":"https://pith.science/paper/J25YCLGL"},"agent_actions":{"view_html":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7","download_json":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7.json","view_paper":"https://pith.science/paper/J25YCLGL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.20246&json=true","fetch_graph":"https://pith.science/api/pith-number/J25YCLGLPWDZ7L7ZQGYV24DXK7/graph.json","fetch_events":"https://pith.science/api/pith-number/J25YCLGLPWDZ7L7ZQGYV24DXK7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7/action/storage_attestation","attest_author":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7/action/author_attestation","sign_citation":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7/action/citation_signature","submit_replication":"https://pith.science/pith/J25YCLGLPWDZ7L7ZQGYV24DXK7/action/replication_record"}},"created_at":"2026-05-21T00:04:22.569236+00:00","updated_at":"2026-05-21T00:04:22.569236+00:00"}