{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GBZ2HTNYI2EX54F5ACI6IDYD2U","short_pith_number":"pith:GBZ2HTNY","schema_version":"1.0","canonical_sha256":"3073a3cdb846897ef0bd0091e40f03d529ef574c6a958009dc5993dc9d399972","source":{"kind":"arxiv","id":"2608.00335","version":1},"attestation_state":"computed","paper":{"title":"RMSWeb: Reflection, Failure-Mode Mining, and Salvage-DS for Web Agent Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Ao Sun, Chengbo Liu, Guichun Hua, Haojun Huang, Lifang Zhou, Pei Tan, Ruijie Yan, Sining Wei, Yingying He, Yining Chen, Yutao Xie","submitted_at":"2026-07-31T22:57:16Z","abstract_excerpt":"Compact web agents can reduce deployment cost, but training them poses challenges in both data collection and post-SFT reinforcement learning (RL). Successful trajectories are expensive to collect and often contain inefficient detours. After supervised fine-tuning (SFT), full trajectory corpora are dominated by routine states; moreover, when group-relative RL is applied to web actions, inadequately designed action-level rewards can yield weak or misleading relative updates, while groups rejected as unsuitable for such updates receive no fallback learning signal. We present RMSWeb, a three-part"},"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":"2608.00335","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-07-31T22:57:16Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"2132cc6d91d24641fc038da029ba44183bb4cc724ec287f91e13851f5f9f0a85","abstract_canon_sha256":"cd0bb9b3c5c22d8199f23ae2c042ceba523d336a7a9134cc5759c4f438f04ab1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T00:35:07.192430Z","signature_b64":"pinJ/GZgE5lQScJyGvw+FojTFoJfTXeVKE9K6j5aKRIy8Vz5784p8XszkqDG5lqe1d9h0EvKRFmzRgGKc5O2Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3073a3cdb846897ef0bd0091e40f03d529ef574c6a958009dc5993dc9d399972","last_reissued_at":"2026-08-04T00:35:07.191008Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T00:35:07.191008Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RMSWeb: Reflection, Failure-Mode Mining, and Salvage-DS for Web Agent Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Ao Sun, Chengbo Liu, Guichun Hua, Haojun Huang, Lifang Zhou, Pei Tan, Ruijie Yan, Sining Wei, Yingying He, Yining Chen, Yutao Xie","submitted_at":"2026-07-31T22:57:16Z","abstract_excerpt":"Compact web agents can reduce deployment cost, but training them poses challenges in both data collection and post-SFT reinforcement learning (RL). Successful trajectories are expensive to collect and often contain inefficient detours. After supervised fine-tuning (SFT), full trajectory corpora are dominated by routine states; moreover, when group-relative RL is applied to web actions, inadequately designed action-level rewards can yield weak or misleading relative updates, while groups rejected as unsuitable for such updates receive no fallback learning signal. We present RMSWeb, a three-part"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.00335","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/2608.00335/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":"2608.00335","created_at":"2026-08-04T00:35:07.192094+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.00335v1","created_at":"2026-08-04T00:35:07.192094+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.00335","created_at":"2026-08-04T00:35:07.192094+00:00"},{"alias_kind":"pith_short_12","alias_value":"GBZ2HTNYI2EX","created_at":"2026-08-04T00:35:07.192094+00:00"},{"alias_kind":"pith_short_16","alias_value":"GBZ2HTNYI2EX54F5","created_at":"2026-08-04T00:35:07.192094+00:00"},{"alias_kind":"pith_short_8","alias_value":"GBZ2HTNY","created_at":"2026-08-04T00:35:07.192094+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/GBZ2HTNYI2EX54F5ACI6IDYD2U","json":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U.json","graph_json":"https://pith.science/api/pith-number/GBZ2HTNYI2EX54F5ACI6IDYD2U/graph.json","events_json":"https://pith.science/api/pith-number/GBZ2HTNYI2EX54F5ACI6IDYD2U/events.json","paper":"https://pith.science/paper/GBZ2HTNY"},"agent_actions":{"view_html":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U","download_json":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U.json","view_paper":"https://pith.science/paper/GBZ2HTNY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.00335&json=true","fetch_graph":"https://pith.science/api/pith-number/GBZ2HTNYI2EX54F5ACI6IDYD2U/graph.json","fetch_events":"https://pith.science/api/pith-number/GBZ2HTNYI2EX54F5ACI6IDYD2U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U/action/storage_attestation","attest_author":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U/action/author_attestation","sign_citation":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U/action/citation_signature","submit_replication":"https://pith.science/pith/GBZ2HTNYI2EX54F5ACI6IDYD2U/action/replication_record"}},"created_at":"2026-08-04T00:35:07.192094+00:00","updated_at":"2026-08-04T00:35:07.192094+00:00"}