{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:XJQ6CVJTQPMWE5P3A44NO4O7PJ","short_pith_number":"pith:XJQ6CVJT","canonical_record":{"source":{"id":"2410.06431","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-09T00:09:15Z","cross_cats_sorted":[],"title_canon_sha256":"3e790a8c80787ba99f36fa3147588a9074e25cd700f2ca8831f217f42fb41a47","abstract_canon_sha256":"6382cec11cecbffa1b5378cbef0ebb22440fa8c3d7e4aa34fd32a6b280c27e9b"},"schema_version":"1.0"},"canonical_sha256":"ba61e1553383d96275fb0738d771df7a62c31586d7647a035db64624f9190216","source":{"kind":"arxiv","id":"2410.06431","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.06431","created_at":"2026-05-17T23:39:18Z"},{"alias_kind":"arxiv_version","alias_value":"2410.06431v5","created_at":"2026-05-17T23:39:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.06431","created_at":"2026-05-17T23:39:18Z"},{"alias_kind":"pith_short_12","alias_value":"XJQ6CVJTQPMW","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"XJQ6CVJTQPMWE5P3","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"XJQ6CVJT","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:XJQ6CVJTQPMWE5P3A44NO4O7PJ","target":"record","payload":{"canonical_record":{"source":{"id":"2410.06431","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-09T00:09:15Z","cross_cats_sorted":[],"title_canon_sha256":"3e790a8c80787ba99f36fa3147588a9074e25cd700f2ca8831f217f42fb41a47","abstract_canon_sha256":"6382cec11cecbffa1b5378cbef0ebb22440fa8c3d7e4aa34fd32a6b280c27e9b"},"schema_version":"1.0"},"canonical_sha256":"ba61e1553383d96275fb0738d771df7a62c31586d7647a035db64624f9190216","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:18.327653Z","signature_b64":"28LyYy/9v+LSiJtAFgISjtdQJJ/8JCaSfgw2S2/JfwiyKfrK17bne6vCb9ouLYltARXzEfOhUeLuO2csfMP7BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba61e1553383d96275fb0738d771df7a62c31586d7647a035db64624f9190216","last_reissued_at":"2026-05-17T23:39:18.326931Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:18.326931Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.06431","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7PDz4qsWI9vX/Uo4S16nfiZHWjOGMCUjLLsnwP1iWNoF521mNvrbZKP1Cyij4tfQqaPLHj8n0XU7bkZPaFW2DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T18:05:38.060786Z"},"content_sha256":"98d2d6b9c35966851a8cb8ac0b747d79dc51605806d89258cb3f132e5f472a3e","schema_version":"1.0","event_id":"sha256:98d2d6b9c35966851a8cb8ac0b747d79dc51605806d89258cb3f132e5f472a3e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:XJQ6CVJTQPMWE5P3A44NO4O7PJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Functional-level Uncertainty Quantification for Calibrated Fine-tuning on LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dongxia Wu, Rose Yu, Ruijia Niu, Yi-An Ma","submitted_at":"2024-10-09T00:09:15Z","abstract_excerpt":"Accurate uncertainty quantification in large language models (LLMs) is essential for reliable confidence estimation, yet fine-tuned LLMs often become overconfident under limited adaptation data. Existing uncertainty methods for PEFT-based LLMs are largely post hoc, estimating uncertainty after fine-tuning rather than improving how adapters specialize to task-specific input-output relationships. We propose Functional-Level Uncertainty Quantification for Calibrated Fine-Tuning (UQ4CT), which calibrates uncertainty over the functional space induced by prompt-dependent mixtures of LoRA experts. UQ"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.06431","kind":"arxiv","version":5},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"2026-05-17T23:39:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"26O/kDCeOpEx3P7mZpXKvf7ms0czcdho8IPr5HImG7+N2uzZQQZZAYx7i/Lf6McHLXUc6NkseXz++4Nh3aOTBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T18:05:38.061471Z"},"content_sha256":"3d1df45dae780bd101507b505ab1779351b5ef2512c8c3951bf3c25c06a5c7ac","schema_version":"1.0","event_id":"sha256:3d1df45dae780bd101507b505ab1779351b5ef2512c8c3951bf3c25c06a5c7ac"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XJQ6CVJTQPMWE5P3A44NO4O7PJ/bundle.json","state_url":"https://pith.science/pith/XJQ6CVJTQPMWE5P3A44NO4O7PJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XJQ6CVJTQPMWE5P3A44NO4O7PJ/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":"2026-05-29T18:05:38Z","links":{"resolver":"https://pith.science/pith/XJQ6CVJTQPMWE5P3A44NO4O7PJ","bundle":"https://pith.science/pith/XJQ6CVJTQPMWE5P3A44NO4O7PJ/bundle.json","state":"https://pith.science/pith/XJQ6CVJTQPMWE5P3A44NO4O7PJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XJQ6CVJTQPMWE5P3A44NO4O7PJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:XJQ6CVJTQPMWE5P3A44NO4O7PJ","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":"6382cec11cecbffa1b5378cbef0ebb22440fa8c3d7e4aa34fd32a6b280c27e9b","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-09T00:09:15Z","title_canon_sha256":"3e790a8c80787ba99f36fa3147588a9074e25cd700f2ca8831f217f42fb41a47"},"schema_version":"1.0","source":{"id":"2410.06431","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.06431","created_at":"2026-05-17T23:39:18Z"},{"alias_kind":"arxiv_version","alias_value":"2410.06431v5","created_at":"2026-05-17T23:39:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.06431","created_at":"2026-05-17T23:39:18Z"},{"alias_kind":"pith_short_12","alias_value":"XJQ6CVJTQPMW","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"XJQ6CVJTQPMWE5P3","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"XJQ6CVJT","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:3d1df45dae780bd101507b505ab1779351b5ef2512c8c3951bf3c25c06a5c7ac","target":"graph","created_at":"2026-05-17T23:39:18Z","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"},"paper":{"abstract_excerpt":"Accurate uncertainty quantification in large language models (LLMs) is essential for reliable confidence estimation, yet fine-tuned LLMs often become overconfident under limited adaptation data. Existing uncertainty methods for PEFT-based LLMs are largely post hoc, estimating uncertainty after fine-tuning rather than improving how adapters specialize to task-specific input-output relationships. We propose Functional-Level Uncertainty Quantification for Calibrated Fine-Tuning (UQ4CT), which calibrates uncertainty over the functional space induced by prompt-dependent mixtures of LoRA experts. UQ","authors_text":"Dongxia Wu, Rose Yu, Ruijia Niu, Yi-An Ma","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-09T00:09:15Z","title":"Functional-level Uncertainty Quantification for Calibrated Fine-tuning on LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.06431","kind":"arxiv","version":5},"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":"sha256:98d2d6b9c35966851a8cb8ac0b747d79dc51605806d89258cb3f132e5f472a3e","target":"record","created_at":"2026-05-17T23:39:18Z","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":"6382cec11cecbffa1b5378cbef0ebb22440fa8c3d7e4aa34fd32a6b280c27e9b","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-09T00:09:15Z","title_canon_sha256":"3e790a8c80787ba99f36fa3147588a9074e25cd700f2ca8831f217f42fb41a47"},"schema_version":"1.0","source":{"id":"2410.06431","kind":"arxiv","version":5}},"canonical_sha256":"ba61e1553383d96275fb0738d771df7a62c31586d7647a035db64624f9190216","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ba61e1553383d96275fb0738d771df7a62c31586d7647a035db64624f9190216","first_computed_at":"2026-05-17T23:39:18.326931Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:18.326931Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"28LyYy/9v+LSiJtAFgISjtdQJJ/8JCaSfgw2S2/JfwiyKfrK17bne6vCb9ouLYltARXzEfOhUeLuO2csfMP7BA==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:18.327653Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.06431","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:98d2d6b9c35966851a8cb8ac0b747d79dc51605806d89258cb3f132e5f472a3e","sha256:3d1df45dae780bd101507b505ab1779351b5ef2512c8c3951bf3c25c06a5c7ac"],"state_sha256":"8d17285ed768ba54077537e8ff1d2f715709bf452588f3e30d69e80e564ab92d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0T/nVzk0NKpZfpkM92raO3nSEAZH8dwVuAwrY9RdfPQ/dPvJhZjBQTf/920Pt6nY1y42yF8tDIv7FgXp/2cXAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-29T18:05:38.065432Z","bundle_sha256":"f002c01ab26d038965af70ddc58455a223d8e95e8ebb0911c126741a61d206b4"}}