{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:2A3DMJ7WYH3FLJFKGMWOAAB7KP","short_pith_number":"pith:2A3DMJ7W","canonical_record":{"source":{"id":"2311.08669","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-15T03:29:02Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"f99df53a3f11b9217734076a8966a0e59531194d5abd0ead601fe6e48d709232","abstract_canon_sha256":"64948134d574134847fce19f5dc47cdb52fe30430a914ed805d7fa18f4e602f5"},"schema_version":"1.0"},"canonical_sha256":"d0363627f6c1f655a4aa332ce0003f53f7095138f676adb71426373af5d8fbdd","source":{"kind":"arxiv","id":"2311.08669","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.08669","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"arxiv_version","alias_value":"2311.08669v2","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.08669","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"pith_short_12","alias_value":"2A3DMJ7WYH3F","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"pith_short_16","alias_value":"2A3DMJ7WYH3FLJFK","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"pith_short_8","alias_value":"2A3DMJ7W","created_at":"2026-07-05T08:07:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:2A3DMJ7WYH3FLJFKGMWOAAB7KP","target":"record","payload":{"canonical_record":{"source":{"id":"2311.08669","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-15T03:29:02Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"f99df53a3f11b9217734076a8966a0e59531194d5abd0ead601fe6e48d709232","abstract_canon_sha256":"64948134d574134847fce19f5dc47cdb52fe30430a914ed805d7fa18f4e602f5"},"schema_version":"1.0"},"canonical_sha256":"d0363627f6c1f655a4aa332ce0003f53f7095138f676adb71426373af5d8fbdd","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:07:59.501919Z","signature_b64":"jQTX9vYDo1GYpIHYPWqidxhIX1aevYk9//N3wAtUrHjpl9PbRnPTuT41y8Bgh4pjA0mhfaCuqMiF8FyjrJbNDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0363627f6c1f655a4aa332ce0003f53f7095138f676adb71426373af5d8fbdd","last_reissued_at":"2026-07-05T08:07:59.501485Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:07:59.501485Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2311.08669","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:07:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"h7dHAzi87q86j5N0z4z2j4SWO+APAswYOkc8MV82hKHL5TF2aD+1WhFlnV//svpM/6uQChgDkQREZPjKWETyAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:19:24.451959Z"},"content_sha256":"667286ec05cc9d7d700f8bd9b19b2385ee35af18c007c20d6ed16b8792fd930f","schema_version":"1.0","event_id":"sha256:667286ec05cc9d7d700f8bd9b19b2385ee35af18c007c20d6ed16b8792fd930f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:2A3DMJ7WYH3FLJFKGMWOAAB7KP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"On the Calibration of Multilingual Question Answering LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Dan Roth, Insup Lee, Soham Dan, Yahan Yang","submitted_at":"2023-11-15T03:29:02Z","abstract_excerpt":"Multilingual pre-trained Large Language Models (LLMs) are incredibly effective at Question Answering (QA), a core task in Natural Language Understanding, achieving high accuracies on several multilingual benchmarks. However, little is known about how well their confidences are calibrated. In this paper, we comprehensively benchmark the calibration of several multilingual LLMs (MLLMs) on a variety of QA tasks. We perform extensive experiments, spanning encoder-only, encoder-decoder, and decoder-only QA models (size varying from 110M to 7B parameters) and diverse languages, including both high- "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.08669","kind":"arxiv","version":2},"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/2311.08669/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":"2026-07-05T08:07:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ug4/BT7HTdHOzr2qOOZASmpr8snt9LAqnrUAsQGsyq1zVqpf/IwmssTEO+Z6fCCXZoT8wZknRpmrClzEq/lzAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:19:24.452765Z"},"content_sha256":"9519bd64bf58bcd762a8f42a109c4aa60b23940e546501cb872b228dc6db4eab","schema_version":"1.0","event_id":"sha256:9519bd64bf58bcd762a8f42a109c4aa60b23940e546501cb872b228dc6db4eab"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2A3DMJ7WYH3FLJFKGMWOAAB7KP/bundle.json","state_url":"https://pith.science/pith/2A3DMJ7WYH3FLJFKGMWOAAB7KP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2A3DMJ7WYH3FLJFKGMWOAAB7KP/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":"2026-08-09T00:19:24Z","links":{"resolver":"https://pith.science/pith/2A3DMJ7WYH3FLJFKGMWOAAB7KP","bundle":"https://pith.science/pith/2A3DMJ7WYH3FLJFKGMWOAAB7KP/bundle.json","state":"https://pith.science/pith/2A3DMJ7WYH3FLJFKGMWOAAB7KP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2A3DMJ7WYH3FLJFKGMWOAAB7KP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:2A3DMJ7WYH3FLJFKGMWOAAB7KP","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":"64948134d574134847fce19f5dc47cdb52fe30430a914ed805d7fa18f4e602f5","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-15T03:29:02Z","title_canon_sha256":"f99df53a3f11b9217734076a8966a0e59531194d5abd0ead601fe6e48d709232"},"schema_version":"1.0","source":{"id":"2311.08669","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.08669","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"arxiv_version","alias_value":"2311.08669v2","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.08669","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"pith_short_12","alias_value":"2A3DMJ7WYH3F","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"pith_short_16","alias_value":"2A3DMJ7WYH3FLJFK","created_at":"2026-07-05T08:07:59Z"},{"alias_kind":"pith_short_8","alias_value":"2A3DMJ7W","created_at":"2026-07-05T08:07:59Z"}],"graph_snapshots":[{"event_id":"sha256:9519bd64bf58bcd762a8f42a109c4aa60b23940e546501cb872b228dc6db4eab","target":"graph","created_at":"2026-07-05T08:07:59Z","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/2311.08669/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multilingual pre-trained Large Language Models (LLMs) are incredibly effective at Question Answering (QA), a core task in Natural Language Understanding, achieving high accuracies on several multilingual benchmarks. However, little is known about how well their confidences are calibrated. In this paper, we comprehensively benchmark the calibration of several multilingual LLMs (MLLMs) on a variety of QA tasks. We perform extensive experiments, spanning encoder-only, encoder-decoder, and decoder-only QA models (size varying from 110M to 7B parameters) and diverse languages, including both high- ","authors_text":"Dan Roth, Insup Lee, Soham Dan, Yahan Yang","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-15T03:29:02Z","title":"On the Calibration of Multilingual Question Answering LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.08669","kind":"arxiv","version":2},"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":"sha256:667286ec05cc9d7d700f8bd9b19b2385ee35af18c007c20d6ed16b8792fd930f","target":"record","created_at":"2026-07-05T08:07:59Z","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":"64948134d574134847fce19f5dc47cdb52fe30430a914ed805d7fa18f4e602f5","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-15T03:29:02Z","title_canon_sha256":"f99df53a3f11b9217734076a8966a0e59531194d5abd0ead601fe6e48d709232"},"schema_version":"1.0","source":{"id":"2311.08669","kind":"arxiv","version":2}},"canonical_sha256":"d0363627f6c1f655a4aa332ce0003f53f7095138f676adb71426373af5d8fbdd","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d0363627f6c1f655a4aa332ce0003f53f7095138f676adb71426373af5d8fbdd","first_computed_at":"2026-07-05T08:07:59.501485Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:07:59.501485Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jQTX9vYDo1GYpIHYPWqidxhIX1aevYk9//N3wAtUrHjpl9PbRnPTuT41y8Bgh4pjA0mhfaCuqMiF8FyjrJbNDw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:07:59.501919Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.08669","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:667286ec05cc9d7d700f8bd9b19b2385ee35af18c007c20d6ed16b8792fd930f","sha256:9519bd64bf58bcd762a8f42a109c4aa60b23940e546501cb872b228dc6db4eab"],"state_sha256":"b134062b7064ad2a472a86298bf539444f2d3ccf3722096f18fc32d4df2899c3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FdEJqsv7+msUsEdzaTzRbDcNeQ60zx0U0/lerJby3MdtcLfrl6ft7vbT+FBqrQTNt4QoGvpIW+WuEJe6GKZdDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T00:19:24.459043Z","bundle_sha256":"3d92d94775b87b2607935a4e820866ab3d0fd776c2e9d819444f8efd3a3a19bd"}}