{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:ZBRL476DFSLKEGW4NQN7SLHK65","short_pith_number":"pith:ZBRL476D","canonical_record":{"source":{"id":"2410.13210","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-17T04:30:46Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"951b99b4bc5e887e4c35d827779e3d877fa5b0d38d6229457de7183029be3c52","abstract_canon_sha256":"1b11101440c027ceabc0a247577f12e44223531ce752f9d34351b6476eedbe53"},"schema_version":"1.0"},"canonical_sha256":"c862be7fc32c96a21adc6c1bf92ceaf7446a9c9c7d747edfb423277cd60c97da","source":{"kind":"arxiv","id":"2410.13210","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.13210","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"arxiv_version","alias_value":"2410.13210v1","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13210","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"pith_short_12","alias_value":"ZBRL476DFSLK","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"pith_short_16","alias_value":"ZBRL476DFSLKEGW4","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"pith_short_8","alias_value":"ZBRL476D","created_at":"2026-07-05T09:21:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:ZBRL476DFSLKEGW4NQN7SLHK65","target":"record","payload":{"canonical_record":{"source":{"id":"2410.13210","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-17T04:30:46Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"951b99b4bc5e887e4c35d827779e3d877fa5b0d38d6229457de7183029be3c52","abstract_canon_sha256":"1b11101440c027ceabc0a247577f12e44223531ce752f9d34351b6476eedbe53"},"schema_version":"1.0"},"canonical_sha256":"c862be7fc32c96a21adc6c1bf92ceaf7446a9c9c7d747edfb423277cd60c97da","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:21:57.140908Z","signature_b64":"QZbIotNqm4S9hbGoAuZV+Nea1qSFr6GDqcNSYZjWrGUinwewSH5m+EubUuXHiHMDMaBPUpuTA49wN1m22e/ODA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c862be7fc32c96a21adc6c1bf92ceaf7446a9c9c7d747edfb423277cd60c97da","last_reissued_at":"2026-07-05T09:21:57.140417Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:21:57.140417Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.13210","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":"2026-07-05T09:21:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"H2wORsroFbe+bMShX58SGeIO7yV9IIr6YjpRO22SL/Vt845j9EnXqlz71LELoKi6whBpVUqptDLwrIjWcIfEAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T16:57:53.004854Z"},"content_sha256":"16020053d7af3dd651bb500b5c158f6f1357bf9b5d421a958fa14497cc62cd8f","schema_version":"1.0","event_id":"sha256:16020053d7af3dd651bb500b5c158f6f1357bf9b5d421a958fa14497cc62cd8f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:ZBRL476DFSLKEGW4NQN7SLHK65","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"FaithBench: A Diverse Hallucination Benchmark for Summarization by Modern LLMs","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amin Ahmad, Chenyu Xu, Erana Wan, Forrest Sheng Bao, Ge Luo, Manveer Singh Tamber, Matthew Gonzales, Miaoran Li, Mike Qi, Ofer Mendelevitch, Renyi Qu, Ruixuan Tu, Suleman Kazi, Vivek Sourabh, Weisi Fan, Yujia Tang","submitted_at":"2024-10-17T04:30:46Z","abstract_excerpt":"Summarization is one of the most common tasks performed by large language models (LLMs), especially in applications like Retrieval-Augmented Generation (RAG). However, existing evaluations of hallucinations in LLM-generated summaries, and evaluations of hallucination detection models both suffer from a lack of diversity and recency in the LLM and LLM families considered. This paper introduces FaithBench, a summarization hallucination benchmark comprising challenging hallucinations made by 10 modern LLMs from 8 different families, with ground truth annotations by human experts. ``Challenging'' "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13210","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/2410.13210/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":"2026-07-05T09:21:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AbaRNrElyh8uDQbq6NK5I4lNlxLsH04Vjwp9DgszGYmGF28yeDy2wdMEMPJkeToLJXQ1KVszp2usTlcjj2O6DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T16:57:53.005225Z"},"content_sha256":"e29492dca1ca57e48adb10e42dabfc27d388d07d41e8fce57119c02860b6f1db","schema_version":"1.0","event_id":"sha256:e29492dca1ca57e48adb10e42dabfc27d388d07d41e8fce57119c02860b6f1db"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZBRL476DFSLKEGW4NQN7SLHK65/bundle.json","state_url":"https://pith.science/pith/ZBRL476DFSLKEGW4NQN7SLHK65/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZBRL476DFSLKEGW4NQN7SLHK65/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":"2026-07-06T16:57:53Z","links":{"resolver":"https://pith.science/pith/ZBRL476DFSLKEGW4NQN7SLHK65","bundle":"https://pith.science/pith/ZBRL476DFSLKEGW4NQN7SLHK65/bundle.json","state":"https://pith.science/pith/ZBRL476DFSLKEGW4NQN7SLHK65/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZBRL476DFSLKEGW4NQN7SLHK65/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:ZBRL476DFSLKEGW4NQN7SLHK65","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":"1b11101440c027ceabc0a247577f12e44223531ce752f9d34351b6476eedbe53","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-17T04:30:46Z","title_canon_sha256":"951b99b4bc5e887e4c35d827779e3d877fa5b0d38d6229457de7183029be3c52"},"schema_version":"1.0","source":{"id":"2410.13210","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.13210","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"arxiv_version","alias_value":"2410.13210v1","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13210","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"pith_short_12","alias_value":"ZBRL476DFSLK","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"pith_short_16","alias_value":"ZBRL476DFSLKEGW4","created_at":"2026-07-05T09:21:57Z"},{"alias_kind":"pith_short_8","alias_value":"ZBRL476D","created_at":"2026-07-05T09:21:57Z"}],"graph_snapshots":[{"event_id":"sha256:e29492dca1ca57e48adb10e42dabfc27d388d07d41e8fce57119c02860b6f1db","target":"graph","created_at":"2026-07-05T09:21:57Z","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/2410.13210/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Summarization is one of the most common tasks performed by large language models (LLMs), especially in applications like Retrieval-Augmented Generation (RAG). However, existing evaluations of hallucinations in LLM-generated summaries, and evaluations of hallucination detection models both suffer from a lack of diversity and recency in the LLM and LLM families considered. This paper introduces FaithBench, a summarization hallucination benchmark comprising challenging hallucinations made by 10 modern LLMs from 8 different families, with ground truth annotations by human experts. ``Challenging'' ","authors_text":"Amin Ahmad, Chenyu Xu, Erana Wan, Forrest Sheng Bao, Ge Luo, Manveer Singh Tamber, Matthew Gonzales, Miaoran Li, Mike Qi, Ofer Mendelevitch, Renyi Qu, Ruixuan Tu, Suleman Kazi, Vivek Sourabh, Weisi Fan, Yujia Tang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-17T04:30:46Z","title":"FaithBench: A Diverse Hallucination Benchmark for Summarization by Modern LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13210","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":"sha256:16020053d7af3dd651bb500b5c158f6f1357bf9b5d421a958fa14497cc62cd8f","target":"record","created_at":"2026-07-05T09:21:57Z","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":"1b11101440c027ceabc0a247577f12e44223531ce752f9d34351b6476eedbe53","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-17T04:30:46Z","title_canon_sha256":"951b99b4bc5e887e4c35d827779e3d877fa5b0d38d6229457de7183029be3c52"},"schema_version":"1.0","source":{"id":"2410.13210","kind":"arxiv","version":1}},"canonical_sha256":"c862be7fc32c96a21adc6c1bf92ceaf7446a9c9c7d747edfb423277cd60c97da","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c862be7fc32c96a21adc6c1bf92ceaf7446a9c9c7d747edfb423277cd60c97da","first_computed_at":"2026-07-05T09:21:57.140417Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:21:57.140417Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"QZbIotNqm4S9hbGoAuZV+Nea1qSFr6GDqcNSYZjWrGUinwewSH5m+EubUuXHiHMDMaBPUpuTA49wN1m22e/ODA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:21:57.140908Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.13210","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:16020053d7af3dd651bb500b5c158f6f1357bf9b5d421a958fa14497cc62cd8f","sha256:e29492dca1ca57e48adb10e42dabfc27d388d07d41e8fce57119c02860b6f1db"],"state_sha256":"c23ff6ad772d89b2e1817fa5d396b951d1df219d2ea2ab08950986f032137ffe"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xiw4JHwdxB4d4kauFj/LxdN/eA5yd+YRBNgixfGNTEgvSy3lyyRuHzIW0LEnOq14wAIQUKtunvXCtIRk3tvrDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T16:57:53.007181Z","bundle_sha256":"5cad159b49c4579bfad95bee5198cef603973fc491bf9cfcf60e96b773868bd5"}}