{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:D4GDQHOEB4HHX75GMJEYGPII5E","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":"91580611236b6338acda16553fc55806761f2460eea60f81e0a170d077da7179","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-23T17:10:20Z","title_canon_sha256":"d8e1af6fbd04c823d9e96f1e6411e9974f47ed792bf9e93954238e5b865c962e"},"schema_version":"1.0","source":{"id":"2605.24667","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.24667","created_at":"2026-05-26T01:03:52Z"},{"alias_kind":"arxiv_version","alias_value":"2605.24667v1","created_at":"2026-05-26T01:03:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.24667","created_at":"2026-05-26T01:03:52Z"},{"alias_kind":"pith_short_12","alias_value":"D4GDQHOEB4HH","created_at":"2026-05-26T01:03:52Z"},{"alias_kind":"pith_short_16","alias_value":"D4GDQHOEB4HHX75G","created_at":"2026-05-26T01:03:52Z"},{"alias_kind":"pith_short_8","alias_value":"D4GDQHOE","created_at":"2026-05-26T01:03:52Z"}],"graph_snapshots":[{"event_id":"sha256:3308bab69658ddc3961eb2beb6afb9294a28617df00a1a6f6783f57cf7f04564","target":"graph","created_at":"2026-05-26T01:03:52Z","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/2605.24667/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Mean cross-entropy is the standard validation metric for language models, but it can fail to track model quality during training. We examine this in two common scenarios. First, in Qwen2.5-1.5B SFT on synthetic fact-learning, we find that mean CE rises substantially after the initial learning phase while held-out fact-recall accuracy remains near its peak. Second, we find that in top-K distillation on TinyStories, decreasing K improves median CE while worsening mean CE; the Top-5 student attains the highest LLM-judge score and crosses below its teacher on median CE, despite having the worst me","authors_text":"Hao Guo, Kevin Shabahang, Rivaan Patil, Simon Dennis","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-23T17:10:20Z","title":"When Mean CE Fails: Median CE Can Better Track Language Model Quality"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.24667","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":"sha256:433e299bde6afcfe1a9e85aa62a73e28c420e650acef5c9b0c78ae0bd6d5f952","target":"record","created_at":"2026-05-26T01:03:52Z","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":"91580611236b6338acda16553fc55806761f2460eea60f81e0a170d077da7179","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-23T17:10:20Z","title_canon_sha256":"d8e1af6fbd04c823d9e96f1e6411e9974f47ed792bf9e93954238e5b865c962e"},"schema_version":"1.0","source":{"id":"2605.24667","kind":"arxiv","version":1}},"canonical_sha256":"1f0c381dc40f0e7bffa66249833d08e9280e6716835a4c86b77c6af8098f6fb5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1f0c381dc40f0e7bffa66249833d08e9280e6716835a4c86b77c6af8098f6fb5","first_computed_at":"2026-05-26T01:03:52.211209Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T01:03:52.211209Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"r8JtsWNPsEpsKQ4Xion5zYxtUNlmSqc7OYtfRAJc+GZljJZlxHh2i+kmD7WFuDnUNgcRvVJ09jCNXIWp2F/UAQ==","signature_status":"signed_v1","signed_at":"2026-05-26T01:03:52.212120Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.24667","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:433e299bde6afcfe1a9e85aa62a73e28c420e650acef5c9b0c78ae0bd6d5f952","sha256:3308bab69658ddc3961eb2beb6afb9294a28617df00a1a6f6783f57cf7f04564"],"state_sha256":"2c4b99e5f7f0a4fdce08dafea24799459a1a91b049d7a6307edbd7269647210a"}