{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:IL2ISXWJBIQCBT3ZXT3QISCMSQ","short_pith_number":"pith:IL2ISXWJ","canonical_record":{"source":{"id":"2503.07631","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-04T19:37:33Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"5391014b2c5889017fb0df054c6115888451b49313c7b3a5da0dd79fbe7479ab","abstract_canon_sha256":"218ef99c63e31b3397c76e670c71f957470a8c217498a3d3f4808bf34b055d44"},"schema_version":"1.0"},"canonical_sha256":"42f4895ec90a2020cf79bcf704484c94207245ff5fd57debba116d769fc7e4b4","source":{"kind":"arxiv","id":"2503.07631","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.07631","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"arxiv_version","alias_value":"2503.07631v3","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.07631","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"pith_short_12","alias_value":"IL2ISXWJBIQC","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"pith_short_16","alias_value":"IL2ISXWJBIQCBT3Z","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"pith_short_8","alias_value":"IL2ISXWJ","created_at":"2026-07-05T11:45:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:IL2ISXWJBIQCBT3ZXT3QISCMSQ","target":"record","payload":{"canonical_record":{"source":{"id":"2503.07631","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-04T19:37:33Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"5391014b2c5889017fb0df054c6115888451b49313c7b3a5da0dd79fbe7479ab","abstract_canon_sha256":"218ef99c63e31b3397c76e670c71f957470a8c217498a3d3f4808bf34b055d44"},"schema_version":"1.0"},"canonical_sha256":"42f4895ec90a2020cf79bcf704484c94207245ff5fd57debba116d769fc7e4b4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:45:32.448510Z","signature_b64":"Ot0gWeUoZI6ca2L2XwHvvAJMz8ALSp2/Y/4jcl3Y94nRdmgSChFLFeS5GnuSd/Jk7vqFO2t+2K/vd5YDpLBIDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"42f4895ec90a2020cf79bcf704484c94207245ff5fd57debba116d769fc7e4b4","last_reissued_at":"2026-07-05T11:45:32.447933Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:45:32.447933Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.07631","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:45:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"s3zju4qJv4ZTykLN72xjAFApa2J5DNKTH6Q9EwyCv658I+HwOS7bgmTJcYRsVaejFvYgpU9lpRd+BTDGi8lgCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T19:25:09.638284Z"},"content_sha256":"1d0e5bc665dd6bcb0f37170a053cfc00f2d50a6a2b953685fae8a8650cd07d7f","schema_version":"1.0","event_id":"sha256:1d0e5bc665dd6bcb0f37170a053cfc00f2d50a6a2b953685fae8a8650cd07d7f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:IL2ISXWJBIQCBT3ZXT3QISCMSQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"OWLViz: An Open-World Benchmark for Visual Question Answering","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Dang Nguyen, Hoang Nguyen, Long Hoang Dang, Thuan Luong, Thuy Nguyen, Viet Dac Lai","submitted_at":"2025-03-04T19:37:33Z","abstract_excerpt":"We present a challenging benchmark for the Open WorLd VISual question answering (OWLViz) task. OWLViz presents concise, unambiguous queries that require integrating multiple capabilities, including visual understanding, web exploration, and specialized tool usage. While humans achieve 69.2% accuracy on these intuitive tasks, even state-of-the-art VLMs struggle, with the best model, Gemini 2.0, achieving only 26.6% accuracy. Current agentic VLMs, which rely on limited vision and vision-language models as tools, perform even worse. This performance gap reveals significant limitations in multimod"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.07631","kind":"arxiv","version":3},"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/2503.07631/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:45:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Efu/tFE2hgAEcNneP0Ok50CX1+2Dh6xWvP2/UnnoTFXlZrMbP2tOYM0Iha/Aogcz30dQknGDRGHGY7BcZ56kAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T19:25:09.638842Z"},"content_sha256":"ae28fae8926c377521b3c7849f9125798fd8c05ddfb4db17c4f7a589a23c0ab2","schema_version":"1.0","event_id":"sha256:ae28fae8926c377521b3c7849f9125798fd8c05ddfb4db17c4f7a589a23c0ab2"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/IL2ISXWJBIQCBT3ZXT3QISCMSQ/bundle.json","state_url":"https://pith.science/pith/IL2ISXWJBIQCBT3ZXT3QISCMSQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/IL2ISXWJBIQCBT3ZXT3QISCMSQ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-05T19:25:09Z","links":{"resolver":"https://pith.science/pith/IL2ISXWJBIQCBT3ZXT3QISCMSQ","bundle":"https://pith.science/pith/IL2ISXWJBIQCBT3ZXT3QISCMSQ/bundle.json","state":"https://pith.science/pith/IL2ISXWJBIQCBT3ZXT3QISCMSQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/IL2ISXWJBIQCBT3ZXT3QISCMSQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:IL2ISXWJBIQCBT3ZXT3QISCMSQ","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":"218ef99c63e31b3397c76e670c71f957470a8c217498a3d3f4808bf34b055d44","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-04T19:37:33Z","title_canon_sha256":"5391014b2c5889017fb0df054c6115888451b49313c7b3a5da0dd79fbe7479ab"},"schema_version":"1.0","source":{"id":"2503.07631","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.07631","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"arxiv_version","alias_value":"2503.07631v3","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.07631","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"pith_short_12","alias_value":"IL2ISXWJBIQC","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"pith_short_16","alias_value":"IL2ISXWJBIQCBT3Z","created_at":"2026-07-05T11:45:32Z"},{"alias_kind":"pith_short_8","alias_value":"IL2ISXWJ","created_at":"2026-07-05T11:45:32Z"}],"graph_snapshots":[{"event_id":"sha256:ae28fae8926c377521b3c7849f9125798fd8c05ddfb4db17c4f7a589a23c0ab2","target":"graph","created_at":"2026-07-05T11:45:32Z","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/2503.07631/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We present a challenging benchmark for the Open WorLd VISual question answering (OWLViz) task. OWLViz presents concise, unambiguous queries that require integrating multiple capabilities, including visual understanding, web exploration, and specialized tool usage. While humans achieve 69.2% accuracy on these intuitive tasks, even state-of-the-art VLMs struggle, with the best model, Gemini 2.0, achieving only 26.6% accuracy. Current agentic VLMs, which rely on limited vision and vision-language models as tools, perform even worse. This performance gap reveals significant limitations in multimod","authors_text":"Dang Nguyen, Hoang Nguyen, Long Hoang Dang, Thuan Luong, Thuy Nguyen, Viet Dac Lai","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-04T19:37:33Z","title":"OWLViz: An Open-World Benchmark for Visual Question Answering"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.07631","kind":"arxiv","version":3},"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":"sha256:1d0e5bc665dd6bcb0f37170a053cfc00f2d50a6a2b953685fae8a8650cd07d7f","target":"record","created_at":"2026-07-05T11:45:32Z","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":"218ef99c63e31b3397c76e670c71f957470a8c217498a3d3f4808bf34b055d44","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-04T19:37:33Z","title_canon_sha256":"5391014b2c5889017fb0df054c6115888451b49313c7b3a5da0dd79fbe7479ab"},"schema_version":"1.0","source":{"id":"2503.07631","kind":"arxiv","version":3}},"canonical_sha256":"42f4895ec90a2020cf79bcf704484c94207245ff5fd57debba116d769fc7e4b4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"42f4895ec90a2020cf79bcf704484c94207245ff5fd57debba116d769fc7e4b4","first_computed_at":"2026-07-05T11:45:32.447933Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:45:32.447933Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Ot0gWeUoZI6ca2L2XwHvvAJMz8ALSp2/Y/4jcl3Y94nRdmgSChFLFeS5GnuSd/Jk7vqFO2t+2K/vd5YDpLBIDg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:45:32.448510Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.07631","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1d0e5bc665dd6bcb0f37170a053cfc00f2d50a6a2b953685fae8a8650cd07d7f","sha256:ae28fae8926c377521b3c7849f9125798fd8c05ddfb4db17c4f7a589a23c0ab2"],"state_sha256":"173fcd5d0d8429356476b000d983a7cb05ccd72c1b087a66b69fba5f3ce6362f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SurmsvW/PVPiEmjk+jn3a+VABGXchuhmdMUNA7og708cRIP8weiTHkEST457zBi0KuZ/qzBPxeJeJMKoqgqtBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T19:25:09.644614Z","bundle_sha256":"4881015d1985e69e4ee544ef5d586b5b0ec248915b7217593d778ad798e1e8ad"}}