{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:Y4KIXWG4UHTRGTNKHBYQSFCJUX","short_pith_number":"pith:Y4KIXWG4","canonical_record":{"source":{"id":"2504.10885","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-15T05:29:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"573d429e1001c0a35c8df26027910c44edd802320a1d54874cbd57dc598d2238","abstract_canon_sha256":"485bf31fda4fdc7848c5f62ad06eebf5b7b1516e3171e365e3d3dadcd262a5a2"},"schema_version":"1.0"},"canonical_sha256":"c7148bd8dca1e7134daa3871091449a5c961804b7cba7cb99580cc663b6360e3","source":{"kind":"arxiv","id":"2504.10885","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.10885","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"arxiv_version","alias_value":"2504.10885v1","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.10885","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"pith_short_12","alias_value":"Y4KIXWG4UHTR","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"pith_short_16","alias_value":"Y4KIXWG4UHTRGTNK","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"pith_short_8","alias_value":"Y4KIXWG4","created_at":"2026-07-05T10:49:25Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:Y4KIXWG4UHTRGTNKHBYQSFCJUX","target":"record","payload":{"canonical_record":{"source":{"id":"2504.10885","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-15T05:29:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"573d429e1001c0a35c8df26027910c44edd802320a1d54874cbd57dc598d2238","abstract_canon_sha256":"485bf31fda4fdc7848c5f62ad06eebf5b7b1516e3171e365e3d3dadcd262a5a2"},"schema_version":"1.0"},"canonical_sha256":"c7148bd8dca1e7134daa3871091449a5c961804b7cba7cb99580cc663b6360e3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:49:25.812947Z","signature_b64":"Y/k4mD8b288eR0DKOvhzsHRdrQHNIYiMn4jx+wwqLTtzmkpD2wYFR3IM4iOlYc7g8jrAtQc55LzhHmsMM6W5Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c7148bd8dca1e7134daa3871091449a5c961804b7cba7cb99580cc663b6360e3","last_reissued_at":"2026-07-05T10:49:25.812446Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:49:25.812446Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2504.10885","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:49:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"h2TFIDUoe27llzsbdxz9CXCDN9gyV7XHbmR/SBWRgFq2fO6mwseles0JBSouGnug1i/urQNXSpA45CMWPdDoBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T06:56:55.218855Z"},"content_sha256":"194e34bd310e49c93e791e7ca08147cd288f9cb99738620a37d8d4a829d1bd45","schema_version":"1.0","event_id":"sha256:194e34bd310e49c93e791e7ca08147cd288f9cb99738620a37d8d4a829d1bd45"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:Y4KIXWG4UHTRGTNKHBYQSFCJUX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"PuzzleBench: A Fully Dynamic Evaluation Framework for Large Multimodal Models on Puzzle Solving","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Chunyi Li, Guangtao Zhai, Xiaohong Liu, Xiongkuo Min, Yuan Tian, Yuze Sun, Zeyu Zhang, Zicheng Zhang, Ziheng Jia, Zijian Chen","submitted_at":"2025-04-15T05:29:31Z","abstract_excerpt":"Large Multimodal Models (LMMs) have demonstrated impressive capabilities across a wide range of multimodal tasks, achieving ever-increasing performance on various evaluation benchmarks. However, existing benchmarks are typically static and often overlap with pre-training datasets, leading to fixed complexity constraints and substantial data contamination issues. Meanwhile, manually annotated datasets are labor-intensive, time-consuming, and subject to human bias and inconsistency, leading to reliability and reproducibility issues. To address these problems, we propose a fully dynamic multimoda"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.10885","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/2504.10885/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":"2026-07-05T10:49:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0n8yQWK1BnXdVCZxMXaT+h0wmt9uj4eMLwH2OzSKZhMvp6twFniA+c7IAzX0S27SNxMZFctZP7udJLlezYGKBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T06:56:55.219223Z"},"content_sha256":"06f60508db69ec900bdc1d24da1b51127023fe630b624f83cee58e3966b508b1","schema_version":"1.0","event_id":"sha256:06f60508db69ec900bdc1d24da1b51127023fe630b624f83cee58e3966b508b1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Y4KIXWG4UHTRGTNKHBYQSFCJUX/bundle.json","state_url":"https://pith.science/pith/Y4KIXWG4UHTRGTNKHBYQSFCJUX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Y4KIXWG4UHTRGTNKHBYQSFCJUX/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":"2026-07-06T06:56:55Z","links":{"resolver":"https://pith.science/pith/Y4KIXWG4UHTRGTNKHBYQSFCJUX","bundle":"https://pith.science/pith/Y4KIXWG4UHTRGTNKHBYQSFCJUX/bundle.json","state":"https://pith.science/pith/Y4KIXWG4UHTRGTNKHBYQSFCJUX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Y4KIXWG4UHTRGTNKHBYQSFCJUX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:Y4KIXWG4UHTRGTNKHBYQSFCJUX","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":"485bf31fda4fdc7848c5f62ad06eebf5b7b1516e3171e365e3d3dadcd262a5a2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-15T05:29:31Z","title_canon_sha256":"573d429e1001c0a35c8df26027910c44edd802320a1d54874cbd57dc598d2238"},"schema_version":"1.0","source":{"id":"2504.10885","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.10885","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"arxiv_version","alias_value":"2504.10885v1","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.10885","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"pith_short_12","alias_value":"Y4KIXWG4UHTR","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"pith_short_16","alias_value":"Y4KIXWG4UHTRGTNK","created_at":"2026-07-05T10:49:25Z"},{"alias_kind":"pith_short_8","alias_value":"Y4KIXWG4","created_at":"2026-07-05T10:49:25Z"}],"graph_snapshots":[{"event_id":"sha256:06f60508db69ec900bdc1d24da1b51127023fe630b624f83cee58e3966b508b1","target":"graph","created_at":"2026-07-05T10:49:25Z","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/2504.10885/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Multimodal Models (LMMs) have demonstrated impressive capabilities across a wide range of multimodal tasks, achieving ever-increasing performance on various evaluation benchmarks. However, existing benchmarks are typically static and often overlap with pre-training datasets, leading to fixed complexity constraints and substantial data contamination issues. Meanwhile, manually annotated datasets are labor-intensive, time-consuming, and subject to human bias and inconsistency, leading to reliability and reproducibility issues. To address these problems, we propose a fully dynamic multimoda","authors_text":"Chunyi Li, Guangtao Zhai, Xiaohong Liu, Xiongkuo Min, Yuan Tian, Yuze Sun, Zeyu Zhang, Zicheng Zhang, Ziheng Jia, Zijian Chen","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-15T05:29:31Z","title":"PuzzleBench: A Fully Dynamic Evaluation Framework for Large Multimodal Models on Puzzle Solving"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.10885","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":"sha256:194e34bd310e49c93e791e7ca08147cd288f9cb99738620a37d8d4a829d1bd45","target":"record","created_at":"2026-07-05T10:49:25Z","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":"485bf31fda4fdc7848c5f62ad06eebf5b7b1516e3171e365e3d3dadcd262a5a2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-15T05:29:31Z","title_canon_sha256":"573d429e1001c0a35c8df26027910c44edd802320a1d54874cbd57dc598d2238"},"schema_version":"1.0","source":{"id":"2504.10885","kind":"arxiv","version":1}},"canonical_sha256":"c7148bd8dca1e7134daa3871091449a5c961804b7cba7cb99580cc663b6360e3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c7148bd8dca1e7134daa3871091449a5c961804b7cba7cb99580cc663b6360e3","first_computed_at":"2026-07-05T10:49:25.812446Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:49:25.812446Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Y/k4mD8b288eR0DKOvhzsHRdrQHNIYiMn4jx+wwqLTtzmkpD2wYFR3IM4iOlYc7g8jrAtQc55LzhHmsMM6W5Aw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:49:25.812947Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.10885","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:194e34bd310e49c93e791e7ca08147cd288f9cb99738620a37d8d4a829d1bd45","sha256:06f60508db69ec900bdc1d24da1b51127023fe630b624f83cee58e3966b508b1"],"state_sha256":"3376796daaba0c39a72b4974430c16e7552afa66bc03a7d27a85b6926949f9b1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0iaqvu1S7GOM4yuMYSetpwrVrfwb4isTzy3kb2hlojUSdB8I7Ls0UApqNitOvKulZJx4cLl/W5cYC9D8NfTtDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T06:56:55.221232Z","bundle_sha256":"7610d5ce8b997a1adb3b993d911806f3375504f2abc02643be1b427b6771b4ea"}}