{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:4DDBIT3ZDIUUTFXVEHSCBXDNC2","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"e84c9afaa46629651a8313b64259ed794a93c5d58f23dc7272e74f36e81cd20e","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-06T23:43:59Z","title_canon_sha256":"94f5f67b767d63c054d97f342b7ff6af5054f78d866c1d83f68fd539c3a0899e"},"schema_version":"1.0","source":{"id":"2412.05467","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.05467","created_at":"2026-07-05T10:21:27Z"},{"alias_kind":"arxiv_version","alias_value":"2412.05467v4","created_at":"2026-07-05T10:21:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.05467","created_at":"2026-07-05T10:21:27Z"},{"alias_kind":"pith_short_12","alias_value":"4DDBIT3ZDIUU","created_at":"2026-07-05T10:21:27Z"},{"alias_kind":"pith_short_16","alias_value":"4DDBIT3ZDIUUTFXV","created_at":"2026-07-05T10:21:27Z"},{"alias_kind":"pith_short_8","alias_value":"4DDBIT3Z","created_at":"2026-07-05T10:21:27Z"}],"graph_snapshots":[{"event_id":"sha256:a8290fd816a8b63131cbf7e4d1bcc8b1a4ae4990ebdc81e6ad21ec88e87dce06","target":"graph","created_at":"2026-07-05T10:21:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2412.05467/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The BrowserGym ecosystem addresses the growing need for efficient evaluation and benchmarking of web agents, particularly those leveraging automation and Large Language Models (LLMs). Many existing benchmarks suffer from fragmentation and inconsistent evaluation methodologies, making it challenging to achieve reliable comparisons and reproducible results. In an earlier work, Drouin et al. (2024) introduced BrowserGym which aims to solve this by providing a unified, gym-like environment with well-defined observation and action spaces, facilitating standardized evaluation across diverse benchmar","authors_text":"Alexandre Drouin, Alexandre Lacoste, Dehan Kong, Frank F. Xu, Graham Neubig, Lawrence Keunho Jang, L\\'eo Boisvert, Massimo Caccia, Maxime Gasse, Megh Thakkar, Nicolas Chapados, Ori Yoran, Quentin Cappart, Rim Assouel, Ruslan Salakhutdinov, Sahar Omidi Shayegan, Siva Reddy, Thibault Le Sellier De Chezelles, Tom Marty, Xing Han L\\`u","cross_cats":["cs.AI","cs.SE"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-06T23:43:59Z","title":"The BrowserGym Ecosystem for Web Agent Research"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.05467","kind":"arxiv","version":4},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:78debd7105706e715179b51624f31911cf807388c02df7ebf06791a1f9562029","target":"record","created_at":"2026-07-05T10:21:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"e84c9afaa46629651a8313b64259ed794a93c5d58f23dc7272e74f36e81cd20e","cross_cats_sorted":["cs.AI","cs.SE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-06T23:43:59Z","title_canon_sha256":"94f5f67b767d63c054d97f342b7ff6af5054f78d866c1d83f68fd539c3a0899e"},"schema_version":"1.0","source":{"id":"2412.05467","kind":"arxiv","version":4}},"canonical_sha256":"e0c6144f791a294996f521e420dc6d16b4834fd37f7aa50cc5112a4319674f83","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e0c6144f791a294996f521e420dc6d16b4834fd37f7aa50cc5112a4319674f83","first_computed_at":"2026-07-05T10:21:27.009069Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:21:27.009069Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4YOwizqttDTkmWxS/GopENduRUQJsr4BZJkfPdgVhFsBZJ4iiYB7/chjk+TtCNWC0ZLKd+LRVSIpvsfzzyeTDg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:21:27.009620Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.05467","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:78debd7105706e715179b51624f31911cf807388c02df7ebf06791a1f9562029","sha256:a8290fd816a8b63131cbf7e4d1bcc8b1a4ae4990ebdc81e6ad21ec88e87dce06"],"state_sha256":"146ce7a89182bceefef2cc85f292a4d102803ad43098635796bdd4fa2e690c77"}