{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:FQ32LH2OFET6F7WFBT36C7DVV6","short_pith_number":"pith:FQ32LH2O","canonical_record":{"source":{"id":"2404.09135","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-14T03:54:00Z","cross_cats_sorted":[],"title_canon_sha256":"1d3c0b8754a216b7fa16abd04d718b3f896dda4aca129002b55cfd8edc10c542","abstract_canon_sha256":"49902c5bc07697e6f7b6e58f0b4f5f0e9e41e7da803fa2581477149b3253897e"},"schema_version":"1.0"},"canonical_sha256":"2c37a59f4e2927e2fec50cf7e17c75af9cc24cdbeef7ef8d895a7db9ef200ee4","source":{"kind":"arxiv","id":"2404.09135","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.09135","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"arxiv_version","alias_value":"2404.09135v1","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.09135","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"pith_short_12","alias_value":"FQ32LH2OFET6","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"pith_short_16","alias_value":"FQ32LH2OFET6F7WF","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"pith_short_8","alias_value":"FQ32LH2O","created_at":"2026-07-05T08:07:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:FQ32LH2OFET6F7WFBT36C7DVV6","target":"record","payload":{"canonical_record":{"source":{"id":"2404.09135","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-14T03:54:00Z","cross_cats_sorted":[],"title_canon_sha256":"1d3c0b8754a216b7fa16abd04d718b3f896dda4aca129002b55cfd8edc10c542","abstract_canon_sha256":"49902c5bc07697e6f7b6e58f0b4f5f0e9e41e7da803fa2581477149b3253897e"},"schema_version":"1.0"},"canonical_sha256":"2c37a59f4e2927e2fec50cf7e17c75af9cc24cdbeef7ef8d895a7db9ef200ee4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:07:45.576500Z","signature_b64":"VszLnGryTQRhxjR3aXDlmjhlxaksp/uhjlzD6Xk0lzypbBwTtZPLNmLlFcKEZeUftJOUA6d1YCARrvLMD8zsCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2c37a59f4e2927e2fec50cf7e17c75af9cc24cdbeef7ef8d895a7db9ef200ee4","last_reissued_at":"2026-07-05T08:07:45.576027Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:07:45.576027Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2404.09135","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":"2026-07-05T08:07:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pGmx9Ey6CVszxy+pA8tqL6CT0LYS2uXJKUVgVYP2I9PM52x7sq2MD9k6hBYx3fUQl2klqLH7K96HshgbWgcjBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T12:41:11.759281Z"},"content_sha256":"84067c319876275cf39ebd9bde60c93c98a8a9f1ad370da35245e5413ebe9fcd","schema_version":"1.0","event_id":"sha256:84067c319876275cf39ebd9bde60c93c98a8a9f1ad370da35245e5413ebe9fcd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:FQ32LH2OFET6F7WFBT36C7DVV6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Unveiling LLM Evaluation Focused on Metrics: Challenges and Solutions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Taojun Hu, Xiao-Hua Zhou","submitted_at":"2024-04-14T03:54:00Z","abstract_excerpt":"Natural Language Processing (NLP) is witnessing a remarkable breakthrough driven by the success of Large Language Models (LLMs). LLMs have gained significant attention across academia and industry for their versatile applications in text generation, question answering, and text summarization. As the landscape of NLP evolves with an increasing number of domain-specific LLMs employing diverse techniques and trained on various corpus, evaluating performance of these models becomes paramount. To quantify the performance, it's crucial to have a comprehensive grasp of existing metrics. Among the eva"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.09135","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/2404.09135/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":"2026-07-05T08:07:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"A7z/T4B+FTmgkXcJy1+9sUBNUkh5vp0jQ+nZKLOYxn47ySx0JYsh5xPV2vk1JghrLhHYhKmBVhLQ0hkCPIgVCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T12:41:11.759798Z"},"content_sha256":"ee850e941d2a8ce8e6353a95d8849c4d1a489a7a99cd95f56ca514636f6dabf8","schema_version":"1.0","event_id":"sha256:ee850e941d2a8ce8e6353a95d8849c4d1a489a7a99cd95f56ca514636f6dabf8"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FQ32LH2OFET6F7WFBT36C7DVV6/bundle.json","state_url":"https://pith.science/pith/FQ32LH2OFET6F7WFBT36C7DVV6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FQ32LH2OFET6F7WFBT36C7DVV6/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":"2026-08-07T12:41:11Z","links":{"resolver":"https://pith.science/pith/FQ32LH2OFET6F7WFBT36C7DVV6","bundle":"https://pith.science/pith/FQ32LH2OFET6F7WFBT36C7DVV6/bundle.json","state":"https://pith.science/pith/FQ32LH2OFET6F7WFBT36C7DVV6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FQ32LH2OFET6F7WFBT36C7DVV6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:FQ32LH2OFET6F7WFBT36C7DVV6","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":"49902c5bc07697e6f7b6e58f0b4f5f0e9e41e7da803fa2581477149b3253897e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-14T03:54:00Z","title_canon_sha256":"1d3c0b8754a216b7fa16abd04d718b3f896dda4aca129002b55cfd8edc10c542"},"schema_version":"1.0","source":{"id":"2404.09135","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.09135","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"arxiv_version","alias_value":"2404.09135v1","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.09135","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"pith_short_12","alias_value":"FQ32LH2OFET6","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"pith_short_16","alias_value":"FQ32LH2OFET6F7WF","created_at":"2026-07-05T08:07:45Z"},{"alias_kind":"pith_short_8","alias_value":"FQ32LH2O","created_at":"2026-07-05T08:07:45Z"}],"graph_snapshots":[{"event_id":"sha256:ee850e941d2a8ce8e6353a95d8849c4d1a489a7a99cd95f56ca514636f6dabf8","target":"graph","created_at":"2026-07-05T08:07:45Z","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/2404.09135/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Natural Language Processing (NLP) is witnessing a remarkable breakthrough driven by the success of Large Language Models (LLMs). LLMs have gained significant attention across academia and industry for their versatile applications in text generation, question answering, and text summarization. As the landscape of NLP evolves with an increasing number of domain-specific LLMs employing diverse techniques and trained on various corpus, evaluating performance of these models becomes paramount. To quantify the performance, it's crucial to have a comprehensive grasp of existing metrics. Among the eva","authors_text":"Taojun Hu, Xiao-Hua Zhou","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-14T03:54:00Z","title":"Unveiling LLM Evaluation Focused on Metrics: Challenges and Solutions"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.09135","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":"sha256:84067c319876275cf39ebd9bde60c93c98a8a9f1ad370da35245e5413ebe9fcd","target":"record","created_at":"2026-07-05T08:07:45Z","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":"49902c5bc07697e6f7b6e58f0b4f5f0e9e41e7da803fa2581477149b3253897e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-14T03:54:00Z","title_canon_sha256":"1d3c0b8754a216b7fa16abd04d718b3f896dda4aca129002b55cfd8edc10c542"},"schema_version":"1.0","source":{"id":"2404.09135","kind":"arxiv","version":1}},"canonical_sha256":"2c37a59f4e2927e2fec50cf7e17c75af9cc24cdbeef7ef8d895a7db9ef200ee4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2c37a59f4e2927e2fec50cf7e17c75af9cc24cdbeef7ef8d895a7db9ef200ee4","first_computed_at":"2026-07-05T08:07:45.576027Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:07:45.576027Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VszLnGryTQRhxjR3aXDlmjhlxaksp/uhjlzD6Xk0lzypbBwTtZPLNmLlFcKEZeUftJOUA6d1YCARrvLMD8zsCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:07:45.576500Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.09135","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:84067c319876275cf39ebd9bde60c93c98a8a9f1ad370da35245e5413ebe9fcd","sha256:ee850e941d2a8ce8e6353a95d8849c4d1a489a7a99cd95f56ca514636f6dabf8"],"state_sha256":"d64d255dba33dd0b4ff1bab31c38ffa92faa20345310506dcaa0e01e95803486"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8ITd+J3XbHc9Ll5D/sLz2EY9pU87xakouePaN96YwjYevj/4nFqHSCV0xCnL1i/JodnOBFVwV5AGuA2+ilJYDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T12:41:11.764694Z","bundle_sha256":"614357a6964fe2a070437bcaa47d155ca6054f8d06a3fe1c54acd86af08f5153"}}