{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:S2PTKWEKFZ72HLWORUXLOZ6X5Q","short_pith_number":"pith:S2PTKWEK","canonical_record":{"source":{"id":"2401.06591","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-12T14:19:23Z","cross_cats_sorted":[],"title_canon_sha256":"7e33c0ecd1ada2ed63beaac77b0097c7b8441e38bfe2137dca88a1f2fd58aee7","abstract_canon_sha256":"b43c1f40e757c6c23933e5a2987c26f1e0f9dde7fa6f7160443489bf7b9b0fe1"},"schema_version":"1.0"},"canonical_sha256":"969f35588a2e7fa3aece8d2eb767d7ec188c378c50183f764205dcb925bdd4c3","source":{"kind":"arxiv","id":"2401.06591","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.06591","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"arxiv_version","alias_value":"2401.06591v1","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.06591","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"pith_short_12","alias_value":"S2PTKWEKFZ72","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"pith_short_16","alias_value":"S2PTKWEKFZ72HLWO","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"pith_short_8","alias_value":"S2PTKWEK","created_at":"2026-07-05T07:33:03Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:S2PTKWEKFZ72HLWORUXLOZ6X5Q","target":"record","payload":{"canonical_record":{"source":{"id":"2401.06591","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-12T14:19:23Z","cross_cats_sorted":[],"title_canon_sha256":"7e33c0ecd1ada2ed63beaac77b0097c7b8441e38bfe2137dca88a1f2fd58aee7","abstract_canon_sha256":"b43c1f40e757c6c23933e5a2987c26f1e0f9dde7fa6f7160443489bf7b9b0fe1"},"schema_version":"1.0"},"canonical_sha256":"969f35588a2e7fa3aece8d2eb767d7ec188c378c50183f764205dcb925bdd4c3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:33:03.208296Z","signature_b64":"AcR7xJNhwvGDfQdZTPL26gea7dT+Ni+tqHuoSyHiRvlIxIYqtvY6MIbIz45gGYSBC6Josq8LKQCKH6DyrUaXBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"969f35588a2e7fa3aece8d2eb767d7ec188c378c50183f764205dcb925bdd4c3","last_reissued_at":"2026-07-05T07:33:03.207859Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:33:03.207859Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2401.06591","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":"2026-07-05T07:33:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"74g7zP5sQZUF1wBaN3Ba2OS+BnpLJGE8H20qY/o7GzwR5rmlUPmGygojOc2O4cKSICDNUuHXWRgcuTtFYQ+nBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T11:52:34.282134Z"},"content_sha256":"6b691ec9ca07ed4a0e673f55d56248e5f6fd9360e9cd8fe16d27f61a447c3b17","schema_version":"1.0","event_id":"sha256:6b691ec9ca07ed4a0e673f55d56248e5f6fd9360e9cd8fe16d27f61a447c3b17"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:S2PTKWEKFZ72HLWORUXLOZ6X5Q","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Prometheus-Vision: Vision-Language Model as a Judge for Fine-Grained Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Geewook Kim, Minjoon Seo, Seongyun Lee, Seungone Kim, Sue Hyun Park","submitted_at":"2024-01-12T14:19:23Z","abstract_excerpt":"Assessing long-form responses generated by Vision-Language Models (VLMs) is challenging. It not only requires checking whether the VLM follows the given instruction but also verifying whether the text output is properly grounded on the given image. Inspired by the recent approach of evaluating LMs with LMs, in this work, we propose to evaluate VLMs with VLMs. For this purpose, we present a new feedback dataset called the Perception Collection, encompassing 15K customized score rubrics that users might care about during assessment. Using the Perception Collection, we train Prometheus-Vision, th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.06591","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/2401.06591/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":"2026-07-05T07:33:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yLKwJ8Q+o65cUBdRfXMTQuRYlDBAQhgT5agtzfC0/bERQBad1+xfjmT1OQZUFmdJANbrl+kgRCSETCF0YkhvDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T11:52:34.282641Z"},"content_sha256":"b09d7849d60cc139587a0b3d5084167968b7ba2267489841d37059d60c3646e3","schema_version":"1.0","event_id":"sha256:b09d7849d60cc139587a0b3d5084167968b7ba2267489841d37059d60c3646e3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/S2PTKWEKFZ72HLWORUXLOZ6X5Q/bundle.json","state_url":"https://pith.science/pith/S2PTKWEKFZ72HLWORUXLOZ6X5Q/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/S2PTKWEKFZ72HLWORUXLOZ6X5Q/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":"2026-08-08T11:52:34Z","links":{"resolver":"https://pith.science/pith/S2PTKWEKFZ72HLWORUXLOZ6X5Q","bundle":"https://pith.science/pith/S2PTKWEKFZ72HLWORUXLOZ6X5Q/bundle.json","state":"https://pith.science/pith/S2PTKWEKFZ72HLWORUXLOZ6X5Q/state.json","well_known_bundle":"https://pith.science/.well-known/pith/S2PTKWEKFZ72HLWORUXLOZ6X5Q/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:S2PTKWEKFZ72HLWORUXLOZ6X5Q","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":"b43c1f40e757c6c23933e5a2987c26f1e0f9dde7fa6f7160443489bf7b9b0fe1","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-12T14:19:23Z","title_canon_sha256":"7e33c0ecd1ada2ed63beaac77b0097c7b8441e38bfe2137dca88a1f2fd58aee7"},"schema_version":"1.0","source":{"id":"2401.06591","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.06591","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"arxiv_version","alias_value":"2401.06591v1","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.06591","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"pith_short_12","alias_value":"S2PTKWEKFZ72","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"pith_short_16","alias_value":"S2PTKWEKFZ72HLWO","created_at":"2026-07-05T07:33:03Z"},{"alias_kind":"pith_short_8","alias_value":"S2PTKWEK","created_at":"2026-07-05T07:33:03Z"}],"graph_snapshots":[{"event_id":"sha256:b09d7849d60cc139587a0b3d5084167968b7ba2267489841d37059d60c3646e3","target":"graph","created_at":"2026-07-05T07:33:03Z","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/2401.06591/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Assessing long-form responses generated by Vision-Language Models (VLMs) is challenging. It not only requires checking whether the VLM follows the given instruction but also verifying whether the text output is properly grounded on the given image. Inspired by the recent approach of evaluating LMs with LMs, in this work, we propose to evaluate VLMs with VLMs. For this purpose, we present a new feedback dataset called the Perception Collection, encompassing 15K customized score rubrics that users might care about during assessment. Using the Perception Collection, we train Prometheus-Vision, th","authors_text":"Geewook Kim, Minjoon Seo, Seongyun Lee, Seungone Kim, Sue Hyun Park","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-12T14:19:23Z","title":"Prometheus-Vision: Vision-Language Model as a Judge for Fine-Grained Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.06591","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":"sha256:6b691ec9ca07ed4a0e673f55d56248e5f6fd9360e9cd8fe16d27f61a447c3b17","target":"record","created_at":"2026-07-05T07:33:03Z","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":"b43c1f40e757c6c23933e5a2987c26f1e0f9dde7fa6f7160443489bf7b9b0fe1","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-12T14:19:23Z","title_canon_sha256":"7e33c0ecd1ada2ed63beaac77b0097c7b8441e38bfe2137dca88a1f2fd58aee7"},"schema_version":"1.0","source":{"id":"2401.06591","kind":"arxiv","version":1}},"canonical_sha256":"969f35588a2e7fa3aece8d2eb767d7ec188c378c50183f764205dcb925bdd4c3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"969f35588a2e7fa3aece8d2eb767d7ec188c378c50183f764205dcb925bdd4c3","first_computed_at":"2026-07-05T07:33:03.207859Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:33:03.207859Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"AcR7xJNhwvGDfQdZTPL26gea7dT+Ni+tqHuoSyHiRvlIxIYqtvY6MIbIz45gGYSBC6Josq8LKQCKH6DyrUaXBg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:33:03.208296Z","signed_message":"canonical_sha256_bytes"},"source_id":"2401.06591","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6b691ec9ca07ed4a0e673f55d56248e5f6fd9360e9cd8fe16d27f61a447c3b17","sha256:b09d7849d60cc139587a0b3d5084167968b7ba2267489841d37059d60c3646e3"],"state_sha256":"d68f4e8553093bb01d616ceffd5d7357771c36e5b1f3f0fa11731983c03bf43c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YHEsJkseV69dE9zUSNOxvb0HVu8RH4Bph5xFIJW2ZxUnqLRbS+TaTUoW1F8zTTDRkyVOgMzXrJQeI0pzRHRuAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T11:52:34.287481Z","bundle_sha256":"26cb96aab2b9f7fc51c76f71be5d4465b6414aa46efe7c25af81db2d999daf79"}}