{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:GDCIDPITZOBLQC6VHYZ3MGQRGW","short_pith_number":"pith:GDCIDPIT","canonical_record":{"source":{"id":"2505.12259","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T06:51:10Z","cross_cats_sorted":[],"title_canon_sha256":"e6341540e4f16001012ab86744853344d0ded173a06f2c9f9798e1666bfdd74d","abstract_canon_sha256":"694f44aef90480b51cbc593c5dd87fa45c7ce476780b6adfabd24ad702b9b993"},"schema_version":"1.0"},"canonical_sha256":"30c481bd13cb82b80bd53e33b61a113589bf8eac44f78d5dcb4030b4cb83de9b","source":{"kind":"arxiv","id":"2505.12259","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12259","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12259v1","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12259","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_12","alias_value":"GDCIDPITZOBL","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_16","alias_value":"GDCIDPITZOBLQC6V","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_8","alias_value":"GDCIDPIT","created_at":"2026-07-05T11:05:01Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:GDCIDPITZOBLQC6VHYZ3MGQRGW","target":"record","payload":{"canonical_record":{"source":{"id":"2505.12259","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T06:51:10Z","cross_cats_sorted":[],"title_canon_sha256":"e6341540e4f16001012ab86744853344d0ded173a06f2c9f9798e1666bfdd74d","abstract_canon_sha256":"694f44aef90480b51cbc593c5dd87fa45c7ce476780b6adfabd24ad702b9b993"},"schema_version":"1.0"},"canonical_sha256":"30c481bd13cb82b80bd53e33b61a113589bf8eac44f78d5dcb4030b4cb83de9b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:05:01.059239Z","signature_b64":"KXHLY8Mh2zxbxMxL95oDtxfc7peOgCugTrFJpKDdUECm+WA4vvo25vxaDFt1y/d4Eh5qAO65Mantki8mv4TwDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"30c481bd13cb82b80bd53e33b61a113589bf8eac44f78d5dcb4030b4cb83de9b","last_reissued_at":"2026-07-05T11:05:01.058703Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:05:01.058703Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.12259","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":"2026-07-05T11:05:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jzks1rUTYoL9X1w5/iGdwkiQVQftn/k+L+2z81BCqxN+UErrcIMX4E0Y4Po0+7xVvlQ2pDwEha3uvZeYVLNWBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T09:35:06.138242Z"},"content_sha256":"40e74ff241d1fc93b1b8f40d21f90853344e7e86f3c8457ce6736d0b99f603ae","schema_version":"1.0","event_id":"sha256:40e74ff241d1fc93b1b8f40d21f90853344e7e86f3c8457ce6736d0b99f603ae"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:GDCIDPITZOBLQC6VHYZ3MGQRGW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Teach2Eval: An Indirect Evaluation Method for LLM by Judging How It Teaches","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chuanjun Ji, Guangnan Ye, Jian Zhang, Siyu Tian, Xiang Liu, Xipeng Qiu, Xutian Chen, Yixin Cao, Yuchen Ni, Yuhang Zhou, Yu He","submitted_at":"2025-05-18T06:51:10Z","abstract_excerpt":"Recent progress in large language models (LLMs) has outpaced the development of effective evaluation methods. Traditional benchmarks rely on task-specific metrics and static datasets, which often suffer from fairness issues, limited scalability, and contamination risks. In this paper, we introduce Teach2Eval, an indirect evaluation framework inspired by the Feynman Technique. Instead of directly testing LLMs on predefined tasks, our method evaluates a model's multiple abilities to teach weaker student models to perform tasks effectively. By converting open-ended tasks into standardized multipl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12259","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/2505.12259/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":"2026-07-05T11:05:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Io0ZBYKjTf03o72usJ9BdOODKuuS7J7UtVR3O1omJ6Ug1SejLNxj76rxiD52T5MS8rwljTayx/ipXcB4G3UFCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T09:35:06.138766Z"},"content_sha256":"c3e0003bf9f60b80404bec5c9f5e18c0ca8d2be0ddc503d5be9af1ecb5a3b34c","schema_version":"1.0","event_id":"sha256:c3e0003bf9f60b80404bec5c9f5e18c0ca8d2be0ddc503d5be9af1ecb5a3b34c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GDCIDPITZOBLQC6VHYZ3MGQRGW/bundle.json","state_url":"https://pith.science/pith/GDCIDPITZOBLQC6VHYZ3MGQRGW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GDCIDPITZOBLQC6VHYZ3MGQRGW/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":"2026-08-14T09:35:06Z","links":{"resolver":"https://pith.science/pith/GDCIDPITZOBLQC6VHYZ3MGQRGW","bundle":"https://pith.science/pith/GDCIDPITZOBLQC6VHYZ3MGQRGW/bundle.json","state":"https://pith.science/pith/GDCIDPITZOBLQC6VHYZ3MGQRGW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GDCIDPITZOBLQC6VHYZ3MGQRGW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:GDCIDPITZOBLQC6VHYZ3MGQRGW","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":"694f44aef90480b51cbc593c5dd87fa45c7ce476780b6adfabd24ad702b9b993","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T06:51:10Z","title_canon_sha256":"e6341540e4f16001012ab86744853344d0ded173a06f2c9f9798e1666bfdd74d"},"schema_version":"1.0","source":{"id":"2505.12259","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12259","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12259v1","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12259","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_12","alias_value":"GDCIDPITZOBL","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_16","alias_value":"GDCIDPITZOBLQC6V","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_8","alias_value":"GDCIDPIT","created_at":"2026-07-05T11:05:01Z"}],"graph_snapshots":[{"event_id":"sha256:c3e0003bf9f60b80404bec5c9f5e18c0ca8d2be0ddc503d5be9af1ecb5a3b34c","target":"graph","created_at":"2026-07-05T11:05:01Z","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/2505.12259/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent progress in large language models (LLMs) has outpaced the development of effective evaluation methods. Traditional benchmarks rely on task-specific metrics and static datasets, which often suffer from fairness issues, limited scalability, and contamination risks. In this paper, we introduce Teach2Eval, an indirect evaluation framework inspired by the Feynman Technique. Instead of directly testing LLMs on predefined tasks, our method evaluates a model's multiple abilities to teach weaker student models to perform tasks effectively. By converting open-ended tasks into standardized multipl","authors_text":"Chuanjun Ji, Guangnan Ye, Jian Zhang, Siyu Tian, Xiang Liu, Xipeng Qiu, Xutian Chen, Yixin Cao, Yuchen Ni, Yuhang Zhou, Yu He","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T06:51:10Z","title":"Teach2Eval: An Indirect Evaluation Method for LLM by Judging How It Teaches"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12259","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":"sha256:40e74ff241d1fc93b1b8f40d21f90853344e7e86f3c8457ce6736d0b99f603ae","target":"record","created_at":"2026-07-05T11:05:01Z","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":"694f44aef90480b51cbc593c5dd87fa45c7ce476780b6adfabd24ad702b9b993","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T06:51:10Z","title_canon_sha256":"e6341540e4f16001012ab86744853344d0ded173a06f2c9f9798e1666bfdd74d"},"schema_version":"1.0","source":{"id":"2505.12259","kind":"arxiv","version":1}},"canonical_sha256":"30c481bd13cb82b80bd53e33b61a113589bf8eac44f78d5dcb4030b4cb83de9b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"30c481bd13cb82b80bd53e33b61a113589bf8eac44f78d5dcb4030b4cb83de9b","first_computed_at":"2026-07-05T11:05:01.058703Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:05:01.058703Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"KXHLY8Mh2zxbxMxL95oDtxfc7peOgCugTrFJpKDdUECm+WA4vvo25vxaDFt1y/d4Eh5qAO65Mantki8mv4TwDg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:05:01.059239Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.12259","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:40e74ff241d1fc93b1b8f40d21f90853344e7e86f3c8457ce6736d0b99f603ae","sha256:c3e0003bf9f60b80404bec5c9f5e18c0ca8d2be0ddc503d5be9af1ecb5a3b34c"],"state_sha256":"5b0d6fee704578e47c879af5702402540827d6a1df8b406fb469fb06583e015c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uhFPJhmUUBWXoM+WqTj7+nbtVfDpk7fyakQ4epu6QcHb660LvWQdPa4fpioIDRUzPkcruRPgVUq0pU9pJ1tBAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T09:35:06.142598Z","bundle_sha256":"79c66a8a7f2d0edd70547dfd6f3dd1a5dbeb055e058062ded967ed90ac0bf47c"}}