{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:UY667D2GHCXQP3PYDV52XE3KPY","short_pith_number":"pith:UY667D2G","canonical_record":{"source":{"id":"2408.01935","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-04T05:24:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1b7caea4249c4aeabffbeb6aa14bd304baa4fc6c30fb4783e22b3cdff135c551","abstract_canon_sha256":"1304c0c52fdb3240d96eff8e5ab54919dd9916e09c98ea03d7fd35f1961dc0b6"},"schema_version":"1.0"},"canonical_sha256":"a63def8f4638af07edf81d7bab936a7e0140ac4aae301fcf97e842b652c3e759","source":{"kind":"arxiv","id":"2408.01935","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.01935","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"arxiv_version","alias_value":"2408.01935v1","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.01935","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"pith_short_12","alias_value":"UY667D2GHCXQ","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"pith_short_16","alias_value":"UY667D2GHCXQP3PY","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"pith_short_8","alias_value":"UY667D2G","created_at":"2026-07-05T08:52:05Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:UY667D2GHCXQP3PYDV52XE3KPY","target":"record","payload":{"canonical_record":{"source":{"id":"2408.01935","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-04T05:24:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1b7caea4249c4aeabffbeb6aa14bd304baa4fc6c30fb4783e22b3cdff135c551","abstract_canon_sha256":"1304c0c52fdb3240d96eff8e5ab54919dd9916e09c98ea03d7fd35f1961dc0b6"},"schema_version":"1.0"},"canonical_sha256":"a63def8f4638af07edf81d7bab936a7e0140ac4aae301fcf97e842b652c3e759","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:52:05.038604Z","signature_b64":"4zVe6QQ/9jFQt27wgx4olFOkoW0q+Fupqs/cqciYKkUFdoB2jztYt96XJIb7cCBHOIcJRs5JcjJCGUEu+XQ/Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a63def8f4638af07edf81d7bab936a7e0140ac4aae301fcf97e842b652c3e759","last_reissued_at":"2026-07-05T08:52:05.038038Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:52:05.038038Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2408.01935","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":"2026-07-05T08:52:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"athg0QDv6xEZOm5sTykSmZsljo+SH8aLNsX9xRWwvHa5Aym4N3WJSFJIPpWYcIicKPrMqcHZVOX9iIwLCSDoBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T02:48:57.437632Z"},"content_sha256":"3ef757e06b3fa26e7c0f8d593890eac7f13ac06b72503bfa465e49324ccce1c6","schema_version":"1.0","event_id":"sha256:3ef757e06b3fa26e7c0f8d593890eac7f13ac06b72503bfa465e49324ccce1c6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:UY667D2GHCXQP3PYDV52XE3KPY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Defining and Evaluating Decision and Composite Risk in Language Models Applied to Natural Language Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ke Shen, Mayank Kejriwal","submitted_at":"2024-08-04T05:24:32Z","abstract_excerpt":"Despite their impressive performance, large language models (LLMs) such as ChatGPT are known to pose important risks. One such set of risks arises from misplaced confidence, whether over-confidence or under-confidence, that the models have in their inference. While the former is well studied, the latter is not, leading to an asymmetry in understanding the comprehensive risk of the model based on misplaced confidence. In this paper, we address this asymmetry by defining two types of risk (decision and composite risk), and proposing an experimental framework consisting of a two-level inference a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.01935","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/2408.01935/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":"2026-07-05T08:52:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yVIbGnZ1awfrqWM65U7ojSSCr3zg1vT/TuuP1jviMoR3lvbJBAclYO5PE1HvpKFJdiGqL6davnezOFwG3Z8+Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T02:48:57.438147Z"},"content_sha256":"47d1cc83052072cfd9b49ad5a0341470637d3fba693b764cdbf37863e5cea242","schema_version":"1.0","event_id":"sha256:47d1cc83052072cfd9b49ad5a0341470637d3fba693b764cdbf37863e5cea242"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/UY667D2GHCXQP3PYDV52XE3KPY/bundle.json","state_url":"https://pith.science/pith/UY667D2GHCXQP3PYDV52XE3KPY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/UY667D2GHCXQP3PYDV52XE3KPY/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":"2026-08-04T02:48:57Z","links":{"resolver":"https://pith.science/pith/UY667D2GHCXQP3PYDV52XE3KPY","bundle":"https://pith.science/pith/UY667D2GHCXQP3PYDV52XE3KPY/bundle.json","state":"https://pith.science/pith/UY667D2GHCXQP3PYDV52XE3KPY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/UY667D2GHCXQP3PYDV52XE3KPY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:UY667D2GHCXQP3PYDV52XE3KPY","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":"1304c0c52fdb3240d96eff8e5ab54919dd9916e09c98ea03d7fd35f1961dc0b6","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-04T05:24:32Z","title_canon_sha256":"1b7caea4249c4aeabffbeb6aa14bd304baa4fc6c30fb4783e22b3cdff135c551"},"schema_version":"1.0","source":{"id":"2408.01935","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.01935","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"arxiv_version","alias_value":"2408.01935v1","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.01935","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"pith_short_12","alias_value":"UY667D2GHCXQ","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"pith_short_16","alias_value":"UY667D2GHCXQP3PY","created_at":"2026-07-05T08:52:05Z"},{"alias_kind":"pith_short_8","alias_value":"UY667D2G","created_at":"2026-07-05T08:52:05Z"}],"graph_snapshots":[{"event_id":"sha256:47d1cc83052072cfd9b49ad5a0341470637d3fba693b764cdbf37863e5cea242","target":"graph","created_at":"2026-07-05T08:52:05Z","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/2408.01935/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite their impressive performance, large language models (LLMs) such as ChatGPT are known to pose important risks. One such set of risks arises from misplaced confidence, whether over-confidence or under-confidence, that the models have in their inference. While the former is well studied, the latter is not, leading to an asymmetry in understanding the comprehensive risk of the model based on misplaced confidence. In this paper, we address this asymmetry by defining two types of risk (decision and composite risk), and proposing an experimental framework consisting of a two-level inference a","authors_text":"Ke Shen, Mayank Kejriwal","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-04T05:24:32Z","title":"Defining and Evaluating Decision and Composite Risk in Language Models Applied to Natural Language Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.01935","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":"sha256:3ef757e06b3fa26e7c0f8d593890eac7f13ac06b72503bfa465e49324ccce1c6","target":"record","created_at":"2026-07-05T08:52:05Z","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":"1304c0c52fdb3240d96eff8e5ab54919dd9916e09c98ea03d7fd35f1961dc0b6","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-04T05:24:32Z","title_canon_sha256":"1b7caea4249c4aeabffbeb6aa14bd304baa4fc6c30fb4783e22b3cdff135c551"},"schema_version":"1.0","source":{"id":"2408.01935","kind":"arxiv","version":1}},"canonical_sha256":"a63def8f4638af07edf81d7bab936a7e0140ac4aae301fcf97e842b652c3e759","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a63def8f4638af07edf81d7bab936a7e0140ac4aae301fcf97e842b652c3e759","first_computed_at":"2026-07-05T08:52:05.038038Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:52:05.038038Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4zVe6QQ/9jFQt27wgx4olFOkoW0q+Fupqs/cqciYKkUFdoB2jztYt96XJIb7cCBHOIcJRs5JcjJCGUEu+XQ/Dw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:52:05.038604Z","signed_message":"canonical_sha256_bytes"},"source_id":"2408.01935","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3ef757e06b3fa26e7c0f8d593890eac7f13ac06b72503bfa465e49324ccce1c6","sha256:47d1cc83052072cfd9b49ad5a0341470637d3fba693b764cdbf37863e5cea242"],"state_sha256":"db7747b74c7e8feb11bb66afb47ae2ed2f30a7f5e1a9e9098933eb9e3e2e3563"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tHys7gsgHRwUrlpslXxceqmlL05dAX8Wi7b1MwpDBiHKYN9dwc109RGd9tL9q7vU6tUQGt6YaYbQoqS+PpVZDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T02:48:57.444019Z","bundle_sha256":"20f91b71b94fa68980d2a1c81fc1516c2e9ffda4eab2fad900da8c0bd01c88aa"}}