{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:FU7EEXD2TWPFMEUGBNZOGWMGG3","short_pith_number":"pith:FU7EEXD2","canonical_record":{"source":{"id":"2503.14941","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-19T07:15:41Z","cross_cats_sorted":[],"title_canon_sha256":"192c39202567e46ec8cbfc666b1fa54f954acb43a8ffdc5051918f14aead3a90","abstract_canon_sha256":"2f26806e716590be976d27df65ed1e974728293bcfdedccd5cc64873ed87b698"},"schema_version":"1.0"},"canonical_sha256":"2d3e425c7a9d9e5612860b72e3598636d6ce03b9dc964bfea8ca4ee3bfb93551","source":{"kind":"arxiv","id":"2503.14941","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.14941","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"arxiv_version","alias_value":"2503.14941v1","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.14941","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"pith_short_12","alias_value":"FU7EEXD2TWPF","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"pith_short_16","alias_value":"FU7EEXD2TWPFMEUG","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"pith_short_8","alias_value":"FU7EEXD2","created_at":"2026-07-05T10:34:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:FU7EEXD2TWPFMEUGBNZOGWMGG3","target":"record","payload":{"canonical_record":{"source":{"id":"2503.14941","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-19T07:15:41Z","cross_cats_sorted":[],"title_canon_sha256":"192c39202567e46ec8cbfc666b1fa54f954acb43a8ffdc5051918f14aead3a90","abstract_canon_sha256":"2f26806e716590be976d27df65ed1e974728293bcfdedccd5cc64873ed87b698"},"schema_version":"1.0"},"canonical_sha256":"2d3e425c7a9d9e5612860b72e3598636d6ce03b9dc964bfea8ca4ee3bfb93551","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:34:28.000587Z","signature_b64":"obtkrFnrGNSFK4cyx9UTz85SPhWTrZt2nAro3mI5jJ2xT5EpG82coy3uIXdpHME1Z8uU1IKhU8I9rBVOjCh5Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2d3e425c7a9d9e5612860b72e3598636d6ce03b9dc964bfea8ca4ee3bfb93551","last_reissued_at":"2026-07-05T10:34:27.999765Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:34:27.999765Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.14941","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":"2026-07-05T10:34:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ALqAK1Tm22VIb3TnPb7RfcordB3fo3q4jzLxUOX6ts1MKeHbPvYiUd3adfwdJMRoLOetdNjYQuE4pHcNmiunAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T03:58:07.110777Z"},"content_sha256":"086453714a68b9370af4643fc89e21317ea261ca5c66c601cd102783993c6476","schema_version":"1.0","event_id":"sha256:086453714a68b9370af4643fc89e21317ea261ca5c66c601cd102783993c6476"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:FU7EEXD2TWPFMEUGBNZOGWMGG3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"UPME: An Unsupervised Peer Review Framework for Multimodal Large Language Model Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jiayi Ye, Li Yuan, Munan Ning, Qihui Zhang, Shuo Yang, Xiao Chen, Yanbo Wang, Yibing Song, Yue Huang, Zheyuan Liu","submitted_at":"2025-03-19T07:15:41Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have emerged to tackle the challenges of Visual Question Answering (VQA), sparking a new research focus on conducting objective evaluations of these models. Existing evaluation methods face limitations due to the significant human workload required to design Q&A pairs for visual images, which inherently restricts the scale and scope of evaluations. Although automated MLLM-as-judge approaches attempt to reduce the human workload through automatic evaluations, they often introduce biases. To address these problems, we propose an Unsupervised Peer review M"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.14941","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/2503.14941/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":"2026-07-05T10:34:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7VDuAyR4UkKjUsMeddoOX8AOPTfPBZzv3Gc0Aa4P6fJPTrx4ujqti267IdNVkDi1VZUD3q3bmUj5eEvmIBgCBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T03:58:07.111302Z"},"content_sha256":"5442ad4eea93bc417b60ba13970eedf28a5bcf02d0509e2518c517552e578b46","schema_version":"1.0","event_id":"sha256:5442ad4eea93bc417b60ba13970eedf28a5bcf02d0509e2518c517552e578b46"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FU7EEXD2TWPFMEUGBNZOGWMGG3/bundle.json","state_url":"https://pith.science/pith/FU7EEXD2TWPFMEUGBNZOGWMGG3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FU7EEXD2TWPFMEUGBNZOGWMGG3/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":"2026-08-10T03:58:07Z","links":{"resolver":"https://pith.science/pith/FU7EEXD2TWPFMEUGBNZOGWMGG3","bundle":"https://pith.science/pith/FU7EEXD2TWPFMEUGBNZOGWMGG3/bundle.json","state":"https://pith.science/pith/FU7EEXD2TWPFMEUGBNZOGWMGG3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FU7EEXD2TWPFMEUGBNZOGWMGG3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:FU7EEXD2TWPFMEUGBNZOGWMGG3","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":"2f26806e716590be976d27df65ed1e974728293bcfdedccd5cc64873ed87b698","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-19T07:15:41Z","title_canon_sha256":"192c39202567e46ec8cbfc666b1fa54f954acb43a8ffdc5051918f14aead3a90"},"schema_version":"1.0","source":{"id":"2503.14941","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.14941","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"arxiv_version","alias_value":"2503.14941v1","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.14941","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"pith_short_12","alias_value":"FU7EEXD2TWPF","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"pith_short_16","alias_value":"FU7EEXD2TWPFMEUG","created_at":"2026-07-05T10:34:27Z"},{"alias_kind":"pith_short_8","alias_value":"FU7EEXD2","created_at":"2026-07-05T10:34:27Z"}],"graph_snapshots":[{"event_id":"sha256:5442ad4eea93bc417b60ba13970eedf28a5bcf02d0509e2518c517552e578b46","target":"graph","created_at":"2026-07-05T10:34:27Z","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/2503.14941/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multimodal Large Language Models (MLLMs) have emerged to tackle the challenges of Visual Question Answering (VQA), sparking a new research focus on conducting objective evaluations of these models. Existing evaluation methods face limitations due to the significant human workload required to design Q&A pairs for visual images, which inherently restricts the scale and scope of evaluations. Although automated MLLM-as-judge approaches attempt to reduce the human workload through automatic evaluations, they often introduce biases. To address these problems, we propose an Unsupervised Peer review M","authors_text":"Jiayi Ye, Li Yuan, Munan Ning, Qihui Zhang, Shuo Yang, Xiao Chen, Yanbo Wang, Yibing Song, Yue Huang, Zheyuan Liu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-19T07:15:41Z","title":"UPME: An Unsupervised Peer Review Framework for Multimodal Large Language Model Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.14941","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":"sha256:086453714a68b9370af4643fc89e21317ea261ca5c66c601cd102783993c6476","target":"record","created_at":"2026-07-05T10:34:27Z","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":"2f26806e716590be976d27df65ed1e974728293bcfdedccd5cc64873ed87b698","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-19T07:15:41Z","title_canon_sha256":"192c39202567e46ec8cbfc666b1fa54f954acb43a8ffdc5051918f14aead3a90"},"schema_version":"1.0","source":{"id":"2503.14941","kind":"arxiv","version":1}},"canonical_sha256":"2d3e425c7a9d9e5612860b72e3598636d6ce03b9dc964bfea8ca4ee3bfb93551","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2d3e425c7a9d9e5612860b72e3598636d6ce03b9dc964bfea8ca4ee3bfb93551","first_computed_at":"2026-07-05T10:34:27.999765Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:34:27.999765Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"obtkrFnrGNSFK4cyx9UTz85SPhWTrZt2nAro3mI5jJ2xT5EpG82coy3uIXdpHME1Z8uU1IKhU8I9rBVOjCh5Dg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:34:28.000587Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.14941","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:086453714a68b9370af4643fc89e21317ea261ca5c66c601cd102783993c6476","sha256:5442ad4eea93bc417b60ba13970eedf28a5bcf02d0509e2518c517552e578b46"],"state_sha256":"7960fe1b6f825ce39af53b8dcc87fa751701c1e64366ffa1c3de574bb6adf020"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"65LPHwH9nYvc47ynZHIl6q9XuWmrip/XIXqlYzGnhSXpiXZhnOyp0NOXKLMRWrNUKWwP/EUjDMwF1+045Dt8Dg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T03:58:07.115433Z","bundle_sha256":"fe9d2715c81e412ae0007eacec470528344ce56dbd71e58b09c6c37638a8165c"}}