{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:ET5XHM36QJ4R4TFDCDWHFSTIDW","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":"64cf4e9342bc2cf2bf631efd1676335b1723ad4e0baa62f24647957094e5efd9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-17T11:25:46Z","title_canon_sha256":"783d8356b3dd044c1cc6836fa9fe981967e7a73cf64f93b0338378195877edd5"},"schema_version":"1.0","source":{"id":"2505.11962","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.11962","created_at":"2026-07-05T11:04:55Z"},{"alias_kind":"arxiv_version","alias_value":"2505.11962v1","created_at":"2026-07-05T11:04:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.11962","created_at":"2026-07-05T11:04:55Z"},{"alias_kind":"pith_short_12","alias_value":"ET5XHM36QJ4R","created_at":"2026-07-05T11:04:55Z"},{"alias_kind":"pith_short_16","alias_value":"ET5XHM36QJ4R4TFD","created_at":"2026-07-05T11:04:55Z"},{"alias_kind":"pith_short_8","alias_value":"ET5XHM36","created_at":"2026-07-05T11:04:55Z"}],"graph_snapshots":[{"event_id":"sha256:36038627b629a50b1b356a91429524fab3b5d04a3a478463ec748e394ddfb522","target":"graph","created_at":"2026-07-05T11:04:55Z","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/2505.11962/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Following instructions in real-world conditions requires the ability to adapt to the world's volatility and entanglement: the environment is dynamic and unpredictable, instructions can be linguistically complex with diverse vocabulary, and the number of possible goals an agent may encounter is vast. Despite extensive research in this area, most studies are conducted in static environments with simple instructions and a limited vocabulary, making it difficult to assess agent performance in more diverse and challenging settings. To address this gap, we introduce CrafText, a benchmark for evaluat","authors_text":"Aleksandr I. Panov, Alexey Skrynnik, Gregory Gorbov, Petr Kuderov, Zoya Volovikova","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-17T11:25:46Z","title":"CrafText Benchmark: Advancing Instruction Following in Complex Multimodal Open-Ended World"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.11962","kind":"arxiv","version":1},"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":"sha256:9dcbc071461955715ac274a5d1ee43aaec0fc4b66daa90ef652de19c8e3d2fe2","target":"record","created_at":"2026-07-05T11:04:55Z","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":"64cf4e9342bc2cf2bf631efd1676335b1723ad4e0baa62f24647957094e5efd9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-17T11:25:46Z","title_canon_sha256":"783d8356b3dd044c1cc6836fa9fe981967e7a73cf64f93b0338378195877edd5"},"schema_version":"1.0","source":{"id":"2505.11962","kind":"arxiv","version":1}},"canonical_sha256":"24fb73b37e82791e4ca310ec72ca681d981a8db4e1b8f1f626f16a909c8e04a1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"24fb73b37e82791e4ca310ec72ca681d981a8db4e1b8f1f626f16a909c8e04a1","first_computed_at":"2026-07-05T11:04:55.524721Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:04:55.524721Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"zKaIAtEQbg8n90R841QH5ce5c8xkQhxNe2JTdnEMPWBtjkNn8jIIJAsmguVd0hnY4UeW2Fzm9El8vZm/i5eXDA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:04:55.525153Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.11962","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9dcbc071461955715ac274a5d1ee43aaec0fc4b66daa90ef652de19c8e3d2fe2","sha256:36038627b629a50b1b356a91429524fab3b5d04a3a478463ec748e394ddfb522"],"state_sha256":"b53d0d6a45ac29290d6420042f5b559e1e691e2a4deee767c1f90b567ff5232a"}