{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:B5L4Q2KUVBBWX2SWBGWUP3UWNB","short_pith_number":"pith:B5L4Q2KU","canonical_record":{"source":{"id":"2409.06851","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-09-10T20:19:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"cce87f29cbd552ab971b8c494bfff4a2269eb99e940354660408d0dec3a38b7e","abstract_canon_sha256":"bc17f3af31f5d85b7c69d10ee9b380f9537155faa96093784cbe2f16304a8d7e"},"schema_version":"1.0"},"canonical_sha256":"0f57c86954a8436bea5609ad47ee96684bb2be23526e16962079936665b4a3bc","source":{"kind":"arxiv","id":"2409.06851","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.06851","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"arxiv_version","alias_value":"2409.06851v3","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.06851","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"pith_short_12","alias_value":"B5L4Q2KUVBBW","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"pith_short_16","alias_value":"B5L4Q2KUVBBWX2SW","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"pith_short_8","alias_value":"B5L4Q2KU","created_at":"2026-07-05T09:20:02Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:B5L4Q2KUVBBWX2SWBGWUP3UWNB","target":"record","payload":{"canonical_record":{"source":{"id":"2409.06851","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-09-10T20:19:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"cce87f29cbd552ab971b8c494bfff4a2269eb99e940354660408d0dec3a38b7e","abstract_canon_sha256":"bc17f3af31f5d85b7c69d10ee9b380f9537155faa96093784cbe2f16304a8d7e"},"schema_version":"1.0"},"canonical_sha256":"0f57c86954a8436bea5609ad47ee96684bb2be23526e16962079936665b4a3bc","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:20:02.577723Z","signature_b64":"cwfdJGu01GofmkbLK9se+SlunMJLkll5l5WEA9mzGTJxqGPJGFXFHSKGqg90CHa6PPH8kVCRHCNdYAY7e0H4CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0f57c86954a8436bea5609ad47ee96684bb2be23526e16962079936665b4a3bc","last_reissued_at":"2026-07-05T09:20:02.577256Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:20:02.577256Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.06851","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":"2026-07-05T09:20:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fNHcebZMyVyjmeGM9vxDXHU5dFH2eJsuVuiR/GgFJIIFlMtg4Vvg3eBhOZ8E2bBiV/IAx2N9u2OaBIi/gLiOBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:01:22.625783Z"},"content_sha256":"c421a9b20062f32be8ac64e5a2d6e960e30c54f9e626c985b85aa2ae935ca2f4","schema_version":"1.0","event_id":"sha256:c421a9b20062f32be8ac64e5a2d6e960e30c54f9e626c985b85aa2ae935ca2f4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:B5L4Q2KUVBBWX2SWBGWUP3UWNB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LIME: Less Is More for MLLM Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Bo Li, Chenghua Lin, Feiteng Fang, Ge Zhang, Haoning Wu, J.H. Liu, Jian Yang, Jiawei Guo, King Zhu, Min Yang, Qianbo Zang, Shawn Gavin, Shian Jia, Shiwen Ni, Siwei Wu, Tuney Zheng, Wenhao Huang, Xiang Yue, Xingwei Qu, Yizhi Li, Zachary Liu","submitted_at":"2024-09-10T20:19:14Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) are evaluated on various benchmarks, such as image captioning, visual question answering, and reasoning. However, many of these benchmarks include overly simple or uninformative samples, complicating the effective distinction of different MLLMs' performance. Furthermore, evaluating models across numerous benchmarks incurs a significant computational burden. To address these issues, we propose LIME (Less Is More for MLLM Evaluation), a refined and efficient benchmark curated through a semi-automated pipeline. This pipeline filters out uninformative sampl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.06851","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/2409.06851/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":"2026-07-05T09:20:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bOOXgCACTJMrp6xtOLVKwy8xYelsIEKSQcbgUEPTViFiaNwmcwtFTe2V5NoxSer+L0hfpFvY0LD8k9TmMt2MBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:01:22.626338Z"},"content_sha256":"efafe62ac0e20a87fb2b3bd6c716dee2f5e8a002c4873d4ea81d121956dc7625","schema_version":"1.0","event_id":"sha256:efafe62ac0e20a87fb2b3bd6c716dee2f5e8a002c4873d4ea81d121956dc7625"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/B5L4Q2KUVBBWX2SWBGWUP3UWNB/bundle.json","state_url":"https://pith.science/pith/B5L4Q2KUVBBWX2SWBGWUP3UWNB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/B5L4Q2KUVBBWX2SWBGWUP3UWNB/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":"2026-08-07T21:01:22Z","links":{"resolver":"https://pith.science/pith/B5L4Q2KUVBBWX2SWBGWUP3UWNB","bundle":"https://pith.science/pith/B5L4Q2KUVBBWX2SWBGWUP3UWNB/bundle.json","state":"https://pith.science/pith/B5L4Q2KUVBBWX2SWBGWUP3UWNB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/B5L4Q2KUVBBWX2SWBGWUP3UWNB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:B5L4Q2KUVBBWX2SWBGWUP3UWNB","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":"bc17f3af31f5d85b7c69d10ee9b380f9537155faa96093784cbe2f16304a8d7e","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-09-10T20:19:14Z","title_canon_sha256":"cce87f29cbd552ab971b8c494bfff4a2269eb99e940354660408d0dec3a38b7e"},"schema_version":"1.0","source":{"id":"2409.06851","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.06851","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"arxiv_version","alias_value":"2409.06851v3","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.06851","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"pith_short_12","alias_value":"B5L4Q2KUVBBW","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"pith_short_16","alias_value":"B5L4Q2KUVBBWX2SW","created_at":"2026-07-05T09:20:02Z"},{"alias_kind":"pith_short_8","alias_value":"B5L4Q2KU","created_at":"2026-07-05T09:20:02Z"}],"graph_snapshots":[{"event_id":"sha256:efafe62ac0e20a87fb2b3bd6c716dee2f5e8a002c4873d4ea81d121956dc7625","target":"graph","created_at":"2026-07-05T09:20:02Z","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/2409.06851/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multimodal Large Language Models (MLLMs) are evaluated on various benchmarks, such as image captioning, visual question answering, and reasoning. However, many of these benchmarks include overly simple or uninformative samples, complicating the effective distinction of different MLLMs' performance. Furthermore, evaluating models across numerous benchmarks incurs a significant computational burden. To address these issues, we propose LIME (Less Is More for MLLM Evaluation), a refined and efficient benchmark curated through a semi-automated pipeline. This pipeline filters out uninformative sampl","authors_text":"Bo Li, Chenghua Lin, Feiteng Fang, Ge Zhang, Haoning Wu, J.H. Liu, Jian Yang, Jiawei Guo, King Zhu, Min Yang, Qianbo Zang, Shawn Gavin, Shian Jia, Shiwen Ni, Siwei Wu, Tuney Zheng, Wenhao Huang, Xiang Yue, Xingwei Qu, Yizhi Li, Zachary Liu","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-09-10T20:19:14Z","title":"LIME: Less Is More for MLLM Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.06851","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:c421a9b20062f32be8ac64e5a2d6e960e30c54f9e626c985b85aa2ae935ca2f4","target":"record","created_at":"2026-07-05T09:20:02Z","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":"bc17f3af31f5d85b7c69d10ee9b380f9537155faa96093784cbe2f16304a8d7e","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-09-10T20:19:14Z","title_canon_sha256":"cce87f29cbd552ab971b8c494bfff4a2269eb99e940354660408d0dec3a38b7e"},"schema_version":"1.0","source":{"id":"2409.06851","kind":"arxiv","version":3}},"canonical_sha256":"0f57c86954a8436bea5609ad47ee96684bb2be23526e16962079936665b4a3bc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0f57c86954a8436bea5609ad47ee96684bb2be23526e16962079936665b4a3bc","first_computed_at":"2026-07-05T09:20:02.577256Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:20:02.577256Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"cwfdJGu01GofmkbLK9se+SlunMJLkll5l5WEA9mzGTJxqGPJGFXFHSKGqg90CHa6PPH8kVCRHCNdYAY7e0H4CA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:20:02.577723Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.06851","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c421a9b20062f32be8ac64e5a2d6e960e30c54f9e626c985b85aa2ae935ca2f4","sha256:efafe62ac0e20a87fb2b3bd6c716dee2f5e8a002c4873d4ea81d121956dc7625"],"state_sha256":"983f77ece8e92e96be15c96c77faae60d0b7cddf0d075b86e44f6a0cde2264bc"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"K27qIvIEak2og74PAl9pzLC4jwKJeqDfbXsVeS6gN74SUdmQLMyTxTSRZeMqxtSHypzTYO8+oYZygSycPirZBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T21:01:22.632083Z","bundle_sha256":"bad7660ffdf9c08c9e40d109710d5b22970e436fac9e7c9feed4c4d210b850a3"}}