{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:R4XXJIEGPTYVIT32QY6VPKRYLS","short_pith_number":"pith:R4XXJIEG","canonical_record":{"source":{"id":"2601.14249","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-20T18:58:10Z","cross_cats_sorted":[],"title_canon_sha256":"b2637a3595c01475d3b4d5444fa2709ba1fd3a4e9e743a9a7a1e2f24391efed0","abstract_canon_sha256":"8908ccfe791646148cc75285abf1995b752c95d5f7e41352ab6c0dd37f76b4a3"},"schema_version":"1.0"},"canonical_sha256":"8f2f74a0867cf1544f7a863d57aa385c93404a161d9db04452c1e317dd1afcea","source":{"kind":"arxiv","id":"2601.14249","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.14249","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"arxiv_version","alias_value":"2601.14249v5","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.14249","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"pith_short_12","alias_value":"R4XXJIEGPTYV","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"pith_short_16","alias_value":"R4XXJIEGPTYVIT32","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"pith_short_8","alias_value":"R4XXJIEG","created_at":"2026-05-26T02:04:03Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:R4XXJIEGPTYVIT32QY6VPKRYLS","target":"record","payload":{"canonical_record":{"source":{"id":"2601.14249","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-20T18:58:10Z","cross_cats_sorted":[],"title_canon_sha256":"b2637a3595c01475d3b4d5444fa2709ba1fd3a4e9e743a9a7a1e2f24391efed0","abstract_canon_sha256":"8908ccfe791646148cc75285abf1995b752c95d5f7e41352ab6c0dd37f76b4a3"},"schema_version":"1.0"},"canonical_sha256":"8f2f74a0867cf1544f7a863d57aa385c93404a161d9db04452c1e317dd1afcea","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:03.267136Z","signature_b64":"1arJvaQbIAQhi4nBegmR4QbJyCR71wVJlvjDqUZNnubPVUYiOd/yXp17kgsUTBaKseych9eEeRQBLpH2miGWAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8f2f74a0867cf1544f7a863d57aa385c93404a161d9db04452c1e317dd1afcea","last_reissued_at":"2026-05-26T02:04:03.266205Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:03.266205Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2601.14249","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":"2026-05-26T02:04:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zmmh8sXgxz4cGFaYFvccMn/LKQOgTNTLZlgFhw17Qz/kwPmKgWeb5hai4m2tPzp7JD577kGpIIBsb2hM3bimBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T02:02:19.202235Z"},"content_sha256":"67aa4cb8c13ef12324c37b9295b96fc80a4dbb882bbb093b8b3ec10afd9182ee","schema_version":"1.0","event_id":"sha256:67aa4cb8c13ef12324c37b9295b96fc80a4dbb882bbb093b8b3ec10afd9182ee"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:R4XXJIEGPTYVIT32QY6VPKRYLS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Which Reasoning Trajectories Teach Students to Reason Better? A Simple Metric of Informative Alignment","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Rank-Surprisal Ratio identifies reasoning trajectories that best improve student model performance","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chiyue Huang, Haijun Lv, Jian Tong, Jun Zhao, Mingqi Wu, Mingyoung Lai, Qipeng Guo, Qi Zhang, Tao Gui, Wanxu Zhao, Xiaoran Fan, Xuanjing Huang, Yicheng Zou, Yuming Yang, Yunhua Zhou, Zhiheng Xi","submitted_at":"2026-01-20T18:58:10Z","abstract_excerpt":"Long chain-of-thought (CoT) trajectories provide rich supervision signals for distilling reasoning from teacher to student LLMs. However, both prior work and our experiments show that trajectories from stronger teachers do not necessarily yield better students, highlighting the importance of data-student suitability in distillation. Existing methods assess suitability primarily through student likelihood, favoring trajectories that align closely with the student model's current behavior but overlooking more informative ones. Addressing this, we propose Rank-Surprisal Ratio (RSR), a simple metr"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across five student models and reasoning trajectories from 11 diverse teachers, RSR strongly correlates with post-training reasoning performance (average Spearman 0.86), consistently outperforming existing metrics.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the balance of low absolute probability and high relative rank under the student model genuinely indicates informativeness rather than an artifact of the tested models, tasks, or trajectory generation methods, and that this correlation will generalize beyond the five students and eleven teachers examined.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Rank-Surprisal Ratio (RSR) correlates strongly (average Spearman 0.86) with post-distillation reasoning gains across five student models and trajectories from eleven teachers, outperforming existing selection metrics.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Rank-Surprisal Ratio identifies reasoning trajectories that best improve student model performance","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7d59564ccc82a42f5e019d23ef79b5d59708acafc8ba123317605e63f4277d27"},"source":{"id":"2601.14249","kind":"arxiv","version":5},"verdict":{"id":"52e45643-fadc-4ef9-a597-bbc0a16aa446","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T12:19:57.928212Z","strongest_claim":"Across five student models and reasoning trajectories from 11 diverse teachers, RSR strongly correlates with post-training reasoning performance (average Spearman 0.86), consistently outperforming existing metrics.","one_line_summary":"Rank-Surprisal Ratio (RSR) correlates strongly (average Spearman 0.86) with post-distillation reasoning gains across five student models and trajectories from eleven teachers, outperforming existing selection metrics.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the balance of low absolute probability and high relative rank under the student model genuinely indicates informativeness rather than an artifact of the tested models, tasks, or trajectory generation methods, and that this correlation will generalize beyond the five students and eleven teachers examined.","pith_extraction_headline":"Rank-Surprisal Ratio identifies reasoning trajectories that best improve student model performance"},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2601.14249/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":"52e45643-fadc-4ef9-a597-bbc0a16aa446"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-26T02:04:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mf0Ok03RDweBwD7deMCjUmcszgCk/dj+sSlJE2tBEIpiOrqamHlj61nil3lnTKBOkuE1zH5J5Ai2jB2xLo9+Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T02:02:19.202919Z"},"content_sha256":"a48fb7251804b2e5fc5f61bdc7d90f80cbb4dcd4587ea90b50c669b758343cf6","schema_version":"1.0","event_id":"sha256:a48fb7251804b2e5fc5f61bdc7d90f80cbb4dcd4587ea90b50c669b758343cf6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/R4XXJIEGPTYVIT32QY6VPKRYLS/bundle.json","state_url":"https://pith.science/pith/R4XXJIEGPTYVIT32QY6VPKRYLS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/R4XXJIEGPTYVIT32QY6VPKRYLS/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":"2026-08-01T02:02:19Z","links":{"resolver":"https://pith.science/pith/R4XXJIEGPTYVIT32QY6VPKRYLS","bundle":"https://pith.science/pith/R4XXJIEGPTYVIT32QY6VPKRYLS/bundle.json","state":"https://pith.science/pith/R4XXJIEGPTYVIT32QY6VPKRYLS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/R4XXJIEGPTYVIT32QY6VPKRYLS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:R4XXJIEGPTYVIT32QY6VPKRYLS","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":"8908ccfe791646148cc75285abf1995b752c95d5f7e41352ab6c0dd37f76b4a3","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-20T18:58:10Z","title_canon_sha256":"b2637a3595c01475d3b4d5444fa2709ba1fd3a4e9e743a9a7a1e2f24391efed0"},"schema_version":"1.0","source":{"id":"2601.14249","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.14249","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"arxiv_version","alias_value":"2601.14249v5","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.14249","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"pith_short_12","alias_value":"R4XXJIEGPTYV","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"pith_short_16","alias_value":"R4XXJIEGPTYVIT32","created_at":"2026-05-26T02:04:03Z"},{"alias_kind":"pith_short_8","alias_value":"R4XXJIEG","created_at":"2026-05-26T02:04:03Z"}],"graph_snapshots":[{"event_id":"sha256:a48fb7251804b2e5fc5f61bdc7d90f80cbb4dcd4587ea90b50c669b758343cf6","target":"graph","created_at":"2026-05-26T02:04:03Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Across five student models and reasoning trajectories from 11 diverse teachers, RSR strongly correlates with post-training reasoning performance (average Spearman 0.86), consistently outperforming existing metrics."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the balance of low absolute probability and high relative rank under the student model genuinely indicates informativeness rather than an artifact of the tested models, tasks, or trajectory generation methods, and that this correlation will generalize beyond the five students and eleven teachers examined."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Rank-Surprisal Ratio (RSR) correlates strongly (average Spearman 0.86) with post-distillation reasoning gains across five student models and trajectories from eleven teachers, outperforming existing selection metrics."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Rank-Surprisal Ratio identifies reasoning trajectories that best improve student model performance"}],"snapshot_sha256":"7d59564ccc82a42f5e019d23ef79b5d59708acafc8ba123317605e63f4277d27"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2601.14249/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Long chain-of-thought (CoT) trajectories provide rich supervision signals for distilling reasoning from teacher to student LLMs. However, both prior work and our experiments show that trajectories from stronger teachers do not necessarily yield better students, highlighting the importance of data-student suitability in distillation. Existing methods assess suitability primarily through student likelihood, favoring trajectories that align closely with the student model's current behavior but overlooking more informative ones. Addressing this, we propose Rank-Surprisal Ratio (RSR), a simple metr","authors_text":"Chiyue Huang, Haijun Lv, Jian Tong, Jun Zhao, Mingqi Wu, Mingyoung Lai, Qipeng Guo, Qi Zhang, Tao Gui, Wanxu Zhao, Xiaoran Fan, Xuanjing Huang, Yicheng Zou, Yuming Yang, Yunhua Zhou, Zhiheng Xi","cross_cats":[],"headline":"Rank-Surprisal Ratio identifies reasoning trajectories that best improve student model performance","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-20T18:58:10Z","title":"Which Reasoning Trajectories Teach Students to Reason Better? A Simple Metric of Informative Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.14249","kind":"arxiv","version":5},"verdict":{"created_at":"2026-05-16T12:19:57.928212Z","id":"52e45643-fadc-4ef9-a597-bbc0a16aa446","model_set":{"reader":"grok-4.3"},"one_line_summary":"Rank-Surprisal Ratio (RSR) correlates strongly (average Spearman 0.86) with post-distillation reasoning gains across five student models and trajectories from eleven teachers, outperforming existing selection metrics.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Rank-Surprisal Ratio identifies reasoning trajectories that best improve student model performance","strongest_claim":"Across five student models and reasoning trajectories from 11 diverse teachers, RSR strongly correlates with post-training reasoning performance (average Spearman 0.86), consistently outperforming existing metrics.","weakest_assumption":"That the balance of low absolute probability and high relative rank under the student model genuinely indicates informativeness rather than an artifact of the tested models, tasks, or trajectory generation methods, and that this correlation will generalize beyond the five students and eleven teachers examined."}},"verdict_id":"52e45643-fadc-4ef9-a597-bbc0a16aa446"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:67aa4cb8c13ef12324c37b9295b96fc80a4dbb882bbb093b8b3ec10afd9182ee","target":"record","created_at":"2026-05-26T02:04:03Z","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":"8908ccfe791646148cc75285abf1995b752c95d5f7e41352ab6c0dd37f76b4a3","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-01-20T18:58:10Z","title_canon_sha256":"b2637a3595c01475d3b4d5444fa2709ba1fd3a4e9e743a9a7a1e2f24391efed0"},"schema_version":"1.0","source":{"id":"2601.14249","kind":"arxiv","version":5}},"canonical_sha256":"8f2f74a0867cf1544f7a863d57aa385c93404a161d9db04452c1e317dd1afcea","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8f2f74a0867cf1544f7a863d57aa385c93404a161d9db04452c1e317dd1afcea","first_computed_at":"2026-05-26T02:04:03.266205Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T02:04:03.266205Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1arJvaQbIAQhi4nBegmR4QbJyCR71wVJlvjDqUZNnubPVUYiOd/yXp17kgsUTBaKseych9eEeRQBLpH2miGWAw==","signature_status":"signed_v1","signed_at":"2026-05-26T02:04:03.267136Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.14249","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:67aa4cb8c13ef12324c37b9295b96fc80a4dbb882bbb093b8b3ec10afd9182ee","sha256:a48fb7251804b2e5fc5f61bdc7d90f80cbb4dcd4587ea90b50c669b758343cf6"],"state_sha256":"cfa1563903ac0933fcbbd01bd7efbcc7c2cc94074e5c64625d339d9f9745884f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"t5ltOEi//8Mq07mv8D9FoZTOVVOIfSPiLvc03xaCA0vzUQBPdmBgVg2Cq0uTOFdf4txsVQcpKRnSQkr/1x3PBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-01T02:02:19.207305Z","bundle_sha256":"13f74840cd0b55456fdc48d293c047c21612f611e881394894a057da1bce1d2c"}}