{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3KACSY7BHCI22JX2BQNHJZCOXZ","short_pith_number":"pith:3KACSY7B","canonical_record":{"source":{"id":"2502.09670","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-12T22:55:43Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7a9779a642a3b805c156810d14172252d7325f3f46b03b4c6724b4317db72c24","abstract_canon_sha256":"f587d25b2c57931a6547c9ff2883ae3e35dbeba30432182ffb9598b8abeae483"},"schema_version":"1.0"},"canonical_sha256":"da802963e13891ad26fa0c1a74e44ebe706d61e6fc2cad44b61cb6bbf78daa97","source":{"kind":"arxiv","id":"2502.09670","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.09670","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"arxiv_version","alias_value":"2502.09670v1","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.09670","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"pith_short_12","alias_value":"3KACSY7BHCI2","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"pith_short_16","alias_value":"3KACSY7BHCI22JX2","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"pith_short_8","alias_value":"3KACSY7B","created_at":"2026-07-05T10:14:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3KACSY7BHCI22JX2BQNHJZCOXZ","target":"record","payload":{"canonical_record":{"source":{"id":"2502.09670","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-12T22:55:43Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7a9779a642a3b805c156810d14172252d7325f3f46b03b4c6724b4317db72c24","abstract_canon_sha256":"f587d25b2c57931a6547c9ff2883ae3e35dbeba30432182ffb9598b8abeae483"},"schema_version":"1.0"},"canonical_sha256":"da802963e13891ad26fa0c1a74e44ebe706d61e6fc2cad44b61cb6bbf78daa97","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:14:10.329235Z","signature_b64":"pohTB5f8J0rAMqM9fTuxMJn4+ELNZy38b06DK0wi5lF2hSXnwx6xq0OXT+sMfWgPkrFi9Qwj9IHJaTPY8UFaBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da802963e13891ad26fa0c1a74e44ebe706d61e6fc2cad44b61cb6bbf78daa97","last_reissued_at":"2026-07-05T10:14:10.328739Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:14:10.328739Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.09670","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":"2026-07-05T10:14:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9/cXU/rd4suZvZDZfnFe/XZESkgSoj/sQa/wFJ76mLyuBeD0wQqCgSpK5/pvIE+2Wr53SZJtJtfNcIXXgEwfBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:50:24.609548Z"},"content_sha256":"55396ccfad6d4049402e24f0ef277238815a806412a71f958fba305158643858","schema_version":"1.0","event_id":"sha256:55396ccfad6d4049402e24f0ef277238815a806412a71f958fba305158643858"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3KACSY7BHCI22JX2BQNHJZCOXZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"The Science of Evaluating Foundation Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Andrew Wen, Jiamu Zhang, Jiayi Yuan, Xia Hu","submitted_at":"2025-02-12T22:55:43Z","abstract_excerpt":"The emergent phenomena of large foundation models have revolutionized natural language processing. However, evaluating these models presents significant challenges due to their size, capabilities, and deployment across diverse applications. Existing literature often focuses on individual aspects, such as benchmark performance or specific tasks, but fails to provide a cohesive process that integrates the nuances of diverse use cases with broader ethical and operational considerations. This work focuses on three key aspects: (1) Formalizing the Evaluation Process by providing a structured framew"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.09670","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/2502.09670/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":"2026-07-05T10:14:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WHJoxO5Aejs/lDe5+y3qZmWE82YSaMrn0ZS0kE4aGq3megnps24UPU4XsgOHn4CALKOCQpFlGK0HAMhwnYtFDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:50:24.610045Z"},"content_sha256":"01bab813434193f64779b6457231eee91d7e421f999eae1194e1642252d3eea0","schema_version":"1.0","event_id":"sha256:01bab813434193f64779b6457231eee91d7e421f999eae1194e1642252d3eea0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3KACSY7BHCI22JX2BQNHJZCOXZ/bundle.json","state_url":"https://pith.science/pith/3KACSY7BHCI22JX2BQNHJZCOXZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3KACSY7BHCI22JX2BQNHJZCOXZ/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":"2026-08-08T22:50:24Z","links":{"resolver":"https://pith.science/pith/3KACSY7BHCI22JX2BQNHJZCOXZ","bundle":"https://pith.science/pith/3KACSY7BHCI22JX2BQNHJZCOXZ/bundle.json","state":"https://pith.science/pith/3KACSY7BHCI22JX2BQNHJZCOXZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3KACSY7BHCI22JX2BQNHJZCOXZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3KACSY7BHCI22JX2BQNHJZCOXZ","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":"f587d25b2c57931a6547c9ff2883ae3e35dbeba30432182ffb9598b8abeae483","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-12T22:55:43Z","title_canon_sha256":"7a9779a642a3b805c156810d14172252d7325f3f46b03b4c6724b4317db72c24"},"schema_version":"1.0","source":{"id":"2502.09670","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.09670","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"arxiv_version","alias_value":"2502.09670v1","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.09670","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"pith_short_12","alias_value":"3KACSY7BHCI2","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"pith_short_16","alias_value":"3KACSY7BHCI22JX2","created_at":"2026-07-05T10:14:10Z"},{"alias_kind":"pith_short_8","alias_value":"3KACSY7B","created_at":"2026-07-05T10:14:10Z"}],"graph_snapshots":[{"event_id":"sha256:01bab813434193f64779b6457231eee91d7e421f999eae1194e1642252d3eea0","target":"graph","created_at":"2026-07-05T10:14:10Z","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/2502.09670/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The emergent phenomena of large foundation models have revolutionized natural language processing. However, evaluating these models presents significant challenges due to their size, capabilities, and deployment across diverse applications. Existing literature often focuses on individual aspects, such as benchmark performance or specific tasks, but fails to provide a cohesive process that integrates the nuances of diverse use cases with broader ethical and operational considerations. This work focuses on three key aspects: (1) Formalizing the Evaluation Process by providing a structured framew","authors_text":"Andrew Wen, Jiamu Zhang, Jiayi Yuan, Xia Hu","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-12T22:55:43Z","title":"The Science of Evaluating Foundation Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.09670","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":"sha256:55396ccfad6d4049402e24f0ef277238815a806412a71f958fba305158643858","target":"record","created_at":"2026-07-05T10:14:10Z","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":"f587d25b2c57931a6547c9ff2883ae3e35dbeba30432182ffb9598b8abeae483","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-12T22:55:43Z","title_canon_sha256":"7a9779a642a3b805c156810d14172252d7325f3f46b03b4c6724b4317db72c24"},"schema_version":"1.0","source":{"id":"2502.09670","kind":"arxiv","version":1}},"canonical_sha256":"da802963e13891ad26fa0c1a74e44ebe706d61e6fc2cad44b61cb6bbf78daa97","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"da802963e13891ad26fa0c1a74e44ebe706d61e6fc2cad44b61cb6bbf78daa97","first_computed_at":"2026-07-05T10:14:10.328739Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:14:10.328739Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"pohTB5f8J0rAMqM9fTuxMJn4+ELNZy38b06DK0wi5lF2hSXnwx6xq0OXT+sMfWgPkrFi9Qwj9IHJaTPY8UFaBw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:14:10.329235Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.09670","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:55396ccfad6d4049402e24f0ef277238815a806412a71f958fba305158643858","sha256:01bab813434193f64779b6457231eee91d7e421f999eae1194e1642252d3eea0"],"state_sha256":"43e992cab5c0eaae4b34a261e2623d1d15a52e1ae5228d0a4666e9590feb7035"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bou5WLTf1iJzY3NcsDeTwtYfQeRxlydPrC8ouY3vF9eQ9WVscKQuofIzCB2ggwiVG96+ns6fPWlcBb4eSlp9Bw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T22:50:24.615261Z","bundle_sha256":"e77dff9cdf8ae8fb7e80b6a9fbdb14cc1755dc0339949bfd9f354694df4fa0da"}}