{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ZD3TXDFRDE2DFYFAYQ7XDWCQCU","short_pith_number":"pith:ZD3TXDFR","canonical_record":{"source":{"id":"2605.17314","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-17T08:12:51Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"f919b4ff0c669173d7777f1632f70a7533f9a1c41be723509d60d1911e3974db","abstract_canon_sha256":"94afb21904af58104ff3468cedd65611840ffd10fce1594d35f8bc6bfb872009"},"schema_version":"1.0"},"canonical_sha256":"c8f73b8cb1193432e0a0c43f71d85015056ad33dca81b2e955f6709740da3c86","source":{"kind":"arxiv","id":"2605.17314","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.17314","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"arxiv_version","alias_value":"2605.17314v1","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17314","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"pith_short_12","alias_value":"ZD3TXDFRDE2D","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"pith_short_16","alias_value":"ZD3TXDFRDE2DFYFA","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"pith_short_8","alias_value":"ZD3TXDFR","created_at":"2026-05-20T00:03:51Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ZD3TXDFRDE2DFYFAYQ7XDWCQCU","target":"record","payload":{"canonical_record":{"source":{"id":"2605.17314","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-17T08:12:51Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"f919b4ff0c669173d7777f1632f70a7533f9a1c41be723509d60d1911e3974db","abstract_canon_sha256":"94afb21904af58104ff3468cedd65611840ffd10fce1594d35f8bc6bfb872009"},"schema_version":"1.0"},"canonical_sha256":"c8f73b8cb1193432e0a0c43f71d85015056ad33dca81b2e955f6709740da3c86","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:03:51.689581Z","signature_b64":"FqC1dlH/u2CRVPeFpauqVWBeSTAJqsCPKpmtU+6jGSi5cQYdfMZZddoJnerx9jbq7knRA1ZpfN4WPhtVbr9BBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c8f73b8cb1193432e0a0c43f71d85015056ad33dca81b2e955f6709740da3c86","last_reissued_at":"2026-05-20T00:03:51.688613Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:03:51.688613Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.17314","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":"2026-05-20T00:03:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kElnpV84MewKfSIndapMnx5iEWAEWasrN73k5cAk7HAScX0F8/KN3w5H8/TppInHa52JLc/FbfipcyYEDCsgAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T19:07:08.741109Z"},"content_sha256":"2005522c2559b2919bd0ee4a74a9728b46067b9c1ed9dc2c2a25b45a5729213f","schema_version":"1.0","event_id":"sha256:2005522c2559b2919bd0ee4a74a9728b46067b9c1ed9dc2c2a25b45a5729213f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ZD3TXDFRDE2DFYFAYQ7XDWCQCU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Weak-to-Strong Elicitation via Mismatched Wrong Drafts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Wei Deng","submitted_at":"2026-05-17T08:12:51Z","abstract_excerpt":"We consider whether off-policy experience from a smaller, weaker model can elicit capability in a stronger learner that on-policy RL fine-tuning (e.g., GRPO) does not reach. We find that injecting mathematically wrong drafts from a smaller but more domain-trained model -- mismatched to the current problem -- into a stronger learner's GRPO context consistently outperforms standard on-policy GRPO on held-out MATH-500 and out-of-distribution AIME 2025/2026. Concretely, we use Mathstral-7B as the learner, Qwen2.5-Math-1.5B as the draft model, 8.8K Level 3--5 MATH problems (with MATH-500 held out),"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17314","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/2605.17314/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-19T22:01:57.783512Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.752665Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"036dbb531853c838a93669c16361e747791775b5965e1e568668f977dab6fdb7"},"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":"2026-05-20T00:03:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XAyh/U8ZMLwspUHskTYd5cpCwohVn5njo2jFzk8McksjYXSejrEYMdOMZKWcGX9J/I1jbP36V937HGVS3M4yBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T19:07:08.741549Z"},"content_sha256":"8b5ba34f4ba6060cf1dc1b51846a25757aba9d99765499697b114b798a69adee","schema_version":"1.0","event_id":"sha256:8b5ba34f4ba6060cf1dc1b51846a25757aba9d99765499697b114b798a69adee"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZD3TXDFRDE2DFYFAYQ7XDWCQCU/bundle.json","state_url":"https://pith.science/pith/ZD3TXDFRDE2DFYFAYQ7XDWCQCU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZD3TXDFRDE2DFYFAYQ7XDWCQCU/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":"2026-05-26T19:07:08Z","links":{"resolver":"https://pith.science/pith/ZD3TXDFRDE2DFYFAYQ7XDWCQCU","bundle":"https://pith.science/pith/ZD3TXDFRDE2DFYFAYQ7XDWCQCU/bundle.json","state":"https://pith.science/pith/ZD3TXDFRDE2DFYFAYQ7XDWCQCU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZD3TXDFRDE2DFYFAYQ7XDWCQCU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ZD3TXDFRDE2DFYFAYQ7XDWCQCU","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":"94afb21904af58104ff3468cedd65611840ffd10fce1594d35f8bc6bfb872009","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-17T08:12:51Z","title_canon_sha256":"f919b4ff0c669173d7777f1632f70a7533f9a1c41be723509d60d1911e3974db"},"schema_version":"1.0","source":{"id":"2605.17314","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.17314","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"arxiv_version","alias_value":"2605.17314v1","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17314","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"pith_short_12","alias_value":"ZD3TXDFRDE2D","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"pith_short_16","alias_value":"ZD3TXDFRDE2DFYFA","created_at":"2026-05-20T00:03:51Z"},{"alias_kind":"pith_short_8","alias_value":"ZD3TXDFR","created_at":"2026-05-20T00:03:51Z"}],"graph_snapshots":[{"event_id":"sha256:8b5ba34f4ba6060cf1dc1b51846a25757aba9d99765499697b114b798a69adee","target":"graph","created_at":"2026-05-20T00:03:51Z","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":[{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-19T22:01:57.783512Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.752665Z","status":"skipped","version":"1.0.0"}],"endpoint":"/pith/2605.17314/integrity.json","findings":[],"snapshot_sha256":"036dbb531853c838a93669c16361e747791775b5965e1e568668f977dab6fdb7","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We consider whether off-policy experience from a smaller, weaker model can elicit capability in a stronger learner that on-policy RL fine-tuning (e.g., GRPO) does not reach. We find that injecting mathematically wrong drafts from a smaller but more domain-trained model -- mismatched to the current problem -- into a stronger learner's GRPO context consistently outperforms standard on-policy GRPO on held-out MATH-500 and out-of-distribution AIME 2025/2026. Concretely, we use Mathstral-7B as the learner, Qwen2.5-Math-1.5B as the draft model, 8.8K Level 3--5 MATH problems (with MATH-500 held out),","authors_text":"Wei Deng","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-17T08:12:51Z","title":"Weak-to-Strong Elicitation via Mismatched Wrong Drafts"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17314","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":"sha256:2005522c2559b2919bd0ee4a74a9728b46067b9c1ed9dc2c2a25b45a5729213f","target":"record","created_at":"2026-05-20T00:03:51Z","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":"94afb21904af58104ff3468cedd65611840ffd10fce1594d35f8bc6bfb872009","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-17T08:12:51Z","title_canon_sha256":"f919b4ff0c669173d7777f1632f70a7533f9a1c41be723509d60d1911e3974db"},"schema_version":"1.0","source":{"id":"2605.17314","kind":"arxiv","version":1}},"canonical_sha256":"c8f73b8cb1193432e0a0c43f71d85015056ad33dca81b2e955f6709740da3c86","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c8f73b8cb1193432e0a0c43f71d85015056ad33dca81b2e955f6709740da3c86","first_computed_at":"2026-05-20T00:03:51.688613Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:03:51.688613Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"FqC1dlH/u2CRVPeFpauqVWBeSTAJqsCPKpmtU+6jGSi5cQYdfMZZddoJnerx9jbq7knRA1ZpfN4WPhtVbr9BBw==","signature_status":"signed_v1","signed_at":"2026-05-20T00:03:51.689581Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.17314","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2005522c2559b2919bd0ee4a74a9728b46067b9c1ed9dc2c2a25b45a5729213f","sha256:8b5ba34f4ba6060cf1dc1b51846a25757aba9d99765499697b114b798a69adee"],"state_sha256":"21120e21a75ed6368ac26810bff75dd0c80e62dc279a024570e00632f3f0919b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"L/eWqVUc39ROnoKI0dc0jxuuo/gxykeTHrZkqYmnKVAVmdtLYmEaPatleSoa4urdQ8sNrQr2IjRljzIJQBLhBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T19:07:08.743988Z","bundle_sha256":"30592252d7149158b8b9c13e0333f48dfa7a76bbba367c5b00bff9702bb56238"}}