{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:OCZA23FAG6DVVEPHF4DQ6LQMMB","short_pith_number":"pith:OCZA23FA","canonical_record":{"source":{"id":"2607.12767","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-14T13:39:11Z","cross_cats_sorted":[],"title_canon_sha256":"b81b077b065274e147da98513cfb8f829b226a06d2915611f8a94865d0c2be5c","abstract_canon_sha256":"d85cd240cb15c3dc028c100f0798a278a6d301aba91c02abdfa8e637abb5e4af"},"schema_version":"1.0"},"canonical_sha256":"70b20d6ca037875a91e72f070f2e0c6053fe4f7b930af69f6f05ef5043d4fae5","source":{"kind":"arxiv","id":"2607.12767","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.12767","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"arxiv_version","alias_value":"2607.12767v1","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.12767","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"pith_short_12","alias_value":"OCZA23FAG6DV","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"pith_short_16","alias_value":"OCZA23FAG6DVVEPH","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"pith_short_8","alias_value":"OCZA23FA","created_at":"2026-07-15T01:22:16Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:OCZA23FAG6DVVEPHF4DQ6LQMMB","target":"record","payload":{"canonical_record":{"source":{"id":"2607.12767","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-14T13:39:11Z","cross_cats_sorted":[],"title_canon_sha256":"b81b077b065274e147da98513cfb8f829b226a06d2915611f8a94865d0c2be5c","abstract_canon_sha256":"d85cd240cb15c3dc028c100f0798a278a6d301aba91c02abdfa8e637abb5e4af"},"schema_version":"1.0"},"canonical_sha256":"70b20d6ca037875a91e72f070f2e0c6053fe4f7b930af69f6f05ef5043d4fae5","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-15T01:22:16.769955Z","signature_b64":"xfieiBUkPqY70q7E7gzAe6wJMtiSV559g/h5/D8/TKwuJOuUjwaPNNQXnzfRSoxc7jS2K9nItw5whhHAb06ZAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"70b20d6ca037875a91e72f070f2e0c6053fe4f7b930af69f6f05ef5043d4fae5","last_reissued_at":"2026-07-15T01:22:16.769137Z","signature_status":"signed_v1","first_computed_at":"2026-07-15T01:22:16.769137Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.12767","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":"2026-07-15T01:22:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KDqb+cjy+PMPtmh0/bEo9r481gGHZCG8MUfIyqeaq7477RDnWLoDgX7ww8btnjkNcR+p8ELIcTAHVLwsO4PcDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T21:43:59.618790Z"},"content_sha256":"b48a7c4507ec61ba3ca499e1bc186d05e0dcc15c1a734be5c1fd17127b60f7f3","schema_version":"1.0","event_id":"sha256:b48a7c4507ec61ba3ca499e1bc186d05e0dcc15c1a734be5c1fd17127b60f7f3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:OCZA23FAG6DVVEPHF4DQ6LQMMB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Accuracy and Normalized Accuracy under Length Bias: Analysis, Guidelines, and a Bayesian Alternative","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Koen Oostermeijer","submitted_at":"2026-07-14T13:39:11Z","abstract_excerpt":"Multiple-choice benchmarks that rank candidate completions by conditional log-probability suffer from a length bias: because log-probabilities sum over tokens, longer answers tend to be penalized relative to shorter ones in practice. A common mitigation is to normalize scores by completion length, but we show empirically that this heuristic frequently over-corrects, introducing a bias toward longer answers instead. We first analyze these scoring rules, characterizing when standard and length-normalized accuracy are appropriate and how their length biases depend on the distribution of completio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.12767","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/2607.12767/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":"2026-07-15T01:22:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TSkwlQsesMVoATOhf1DQIveQVt7wFSPWFO+GRvK/LygdXDqTMpiF4rLac0rfZ4BOu4K4xNRVXT9wAK+GjcWHCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T21:43:59.619705Z"},"content_sha256":"453a5ca06e37781207617cd666445e79ad3a417368e78dfd3a679c09b59ddab1","schema_version":"1.0","event_id":"sha256:453a5ca06e37781207617cd666445e79ad3a417368e78dfd3a679c09b59ddab1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OCZA23FAG6DVVEPHF4DQ6LQMMB/bundle.json","state_url":"https://pith.science/pith/OCZA23FAG6DVVEPHF4DQ6LQMMB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OCZA23FAG6DVVEPHF4DQ6LQMMB/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":"2026-08-03T21:43:59Z","links":{"resolver":"https://pith.science/pith/OCZA23FAG6DVVEPHF4DQ6LQMMB","bundle":"https://pith.science/pith/OCZA23FAG6DVVEPHF4DQ6LQMMB/bundle.json","state":"https://pith.science/pith/OCZA23FAG6DVVEPHF4DQ6LQMMB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OCZA23FAG6DVVEPHF4DQ6LQMMB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:OCZA23FAG6DVVEPHF4DQ6LQMMB","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":"d85cd240cb15c3dc028c100f0798a278a6d301aba91c02abdfa8e637abb5e4af","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-14T13:39:11Z","title_canon_sha256":"b81b077b065274e147da98513cfb8f829b226a06d2915611f8a94865d0c2be5c"},"schema_version":"1.0","source":{"id":"2607.12767","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.12767","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"arxiv_version","alias_value":"2607.12767v1","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.12767","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"pith_short_12","alias_value":"OCZA23FAG6DV","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"pith_short_16","alias_value":"OCZA23FAG6DVVEPH","created_at":"2026-07-15T01:22:16Z"},{"alias_kind":"pith_short_8","alias_value":"OCZA23FA","created_at":"2026-07-15T01:22:16Z"}],"graph_snapshots":[{"event_id":"sha256:453a5ca06e37781207617cd666445e79ad3a417368e78dfd3a679c09b59ddab1","target":"graph","created_at":"2026-07-15T01:22:16Z","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/2607.12767/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multiple-choice benchmarks that rank candidate completions by conditional log-probability suffer from a length bias: because log-probabilities sum over tokens, longer answers tend to be penalized relative to shorter ones in practice. A common mitigation is to normalize scores by completion length, but we show empirically that this heuristic frequently over-corrects, introducing a bias toward longer answers instead. We first analyze these scoring rules, characterizing when standard and length-normalized accuracy are appropriate and how their length biases depend on the distribution of completio","authors_text":"Koen Oostermeijer","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-14T13:39:11Z","title":"Accuracy and Normalized Accuracy under Length Bias: Analysis, Guidelines, and a Bayesian Alternative"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.12767","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:b48a7c4507ec61ba3ca499e1bc186d05e0dcc15c1a734be5c1fd17127b60f7f3","target":"record","created_at":"2026-07-15T01:22:16Z","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":"d85cd240cb15c3dc028c100f0798a278a6d301aba91c02abdfa8e637abb5e4af","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-14T13:39:11Z","title_canon_sha256":"b81b077b065274e147da98513cfb8f829b226a06d2915611f8a94865d0c2be5c"},"schema_version":"1.0","source":{"id":"2607.12767","kind":"arxiv","version":1}},"canonical_sha256":"70b20d6ca037875a91e72f070f2e0c6053fe4f7b930af69f6f05ef5043d4fae5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"70b20d6ca037875a91e72f070f2e0c6053fe4f7b930af69f6f05ef5043d4fae5","first_computed_at":"2026-07-15T01:22:16.769137Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-15T01:22:16.769137Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"xfieiBUkPqY70q7E7gzAe6wJMtiSV559g/h5/D8/TKwuJOuUjwaPNNQXnzfRSoxc7jS2K9nItw5whhHAb06ZAg==","signature_status":"signed_v1","signed_at":"2026-07-15T01:22:16.769955Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.12767","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b48a7c4507ec61ba3ca499e1bc186d05e0dcc15c1a734be5c1fd17127b60f7f3","sha256:453a5ca06e37781207617cd666445e79ad3a417368e78dfd3a679c09b59ddab1"],"state_sha256":"280e3b6fd3d1546f901c4c2bfc9992bcdd2b7e5609e729f385be29af8b3b6f6c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cRoe/aAihxUCVGS78l69REw0ewV/f3OCgLtv2nAXr7/CARB/AVyuvNVOcPa0sjhD7G+HCQcAddAD+k1dUp8KDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T21:43:59.626378Z","bundle_sha256":"3aa410223e6b794067250a3b0c9b230e1dbd459f9f101df10eb6d1987cc640cf"}}