{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:SRJEQMARHX3HZREMR63C3CAPY3","short_pith_number":"pith:SRJEQMAR","canonical_record":{"source":{"id":"2502.19557","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T20:50:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7914667cb52bbe8399bf9b06c6f6602210d329337de07f9f49d84a12270c133f","abstract_canon_sha256":"eb8eb4bac67a4a338eb4a7da4c7a77aadd8a5a5fb0cdc03d0ba28b68d98e6fc3"},"schema_version":"1.0"},"canonical_sha256":"94524830113df67cc48c8fb62d880fc6fdcfd90b404ef5adbef4f745724095e9","source":{"kind":"arxiv","id":"2502.19557","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.19557","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"arxiv_version","alias_value":"2502.19557v1","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.19557","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"pith_short_12","alias_value":"SRJEQMARHX3H","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"pith_short_16","alias_value":"SRJEQMARHX3HZREM","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"pith_short_8","alias_value":"SRJEQMAR","created_at":"2026-07-05T10:20:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:SRJEQMARHX3HZREMR63C3CAPY3","target":"record","payload":{"canonical_record":{"source":{"id":"2502.19557","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T20:50:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7914667cb52bbe8399bf9b06c6f6602210d329337de07f9f49d84a12270c133f","abstract_canon_sha256":"eb8eb4bac67a4a338eb4a7da4c7a77aadd8a5a5fb0cdc03d0ba28b68d98e6fc3"},"schema_version":"1.0"},"canonical_sha256":"94524830113df67cc48c8fb62d880fc6fdcfd90b404ef5adbef4f745724095e9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:20:48.674896Z","signature_b64":"RibY8iONNyUwKIC0SNLZFMTEpW0S0qWVJARZa6H7Nwdvwdo3QfOx6RrYSRP3X+fZSFo6A9u/5W67+YHiMFOWCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"94524830113df67cc48c8fb62d880fc6fdcfd90b404ef5adbef4f745724095e9","last_reissued_at":"2026-07-05T10:20:48.674387Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:20:48.674387Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.19557","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":"2026-07-05T10:20:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SiNHLB9LMdWQzwR/n4boDby0fsmWKMa2oZLicA10JJpSrPhkw1O423dchfmqhN5mBvFPhyIZL5ks+y1uaxisCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T11:38:57.610241Z"},"content_sha256":"7c473f55ffcec25b4c9a7196feb184ad8fddb89dd73f2973cc2e8fc1bada983d","schema_version":"1.0","event_id":"sha256:7c473f55ffcec25b4c9a7196feb184ad8fddb89dd73f2973cc2e8fc1bada983d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:SRJEQMARHX3HZREMR63C3CAPY3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Distill Not Only Data but Also Rewards: Can Smaller Language Models Surpass Larger Ones?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chenghua Huang, Dongmei Zhang, Jun Wang, Lu Wang, Meng Fang, Mykola Pechenizkiy, Qingwei Lin, Qi Zhang, Saravan Rajmohan, Yali Du, Yudi Zhang","submitted_at":"2025-02-26T20:50:11Z","abstract_excerpt":"Distilling large language models (LLMs) typically involves transferring the teacher model's responses through supervised fine-tuning (SFT). However, this approach neglects the potential to distill both data (output content) and reward signals (quality evaluations). Extracting reliable reward signals directly from teacher models is challenging, as LLMs are optimized for generation rather than evaluation, often resulting in biased or inconsistent assessments. To address this limitation, we propose a novel distillation pipeline that transfers both responses and rewards. Our method generates pseud"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.19557","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/2502.19557/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":"2026-07-05T10:20:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hgUBdm+Wzcvd259ggCMxC6ArxaH4IPCKUIh8TqWxU78gUgY9MZew21PHCq22GI9m8fZ+tX0NxFWvpmJhmyeMDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T11:38:57.610634Z"},"content_sha256":"ffd01babf75cb78f58e22fb307d7796e7ea16055fe4a10829e5152e7a789eaa1","schema_version":"1.0","event_id":"sha256:ffd01babf75cb78f58e22fb307d7796e7ea16055fe4a10829e5152e7a789eaa1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SRJEQMARHX3HZREMR63C3CAPY3/bundle.json","state_url":"https://pith.science/pith/SRJEQMARHX3HZREMR63C3CAPY3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SRJEQMARHX3HZREMR63C3CAPY3/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":"2026-07-07T11:38:57Z","links":{"resolver":"https://pith.science/pith/SRJEQMARHX3HZREMR63C3CAPY3","bundle":"https://pith.science/pith/SRJEQMARHX3HZREMR63C3CAPY3/bundle.json","state":"https://pith.science/pith/SRJEQMARHX3HZREMR63C3CAPY3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SRJEQMARHX3HZREMR63C3CAPY3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:SRJEQMARHX3HZREMR63C3CAPY3","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":"eb8eb4bac67a4a338eb4a7da4c7a77aadd8a5a5fb0cdc03d0ba28b68d98e6fc3","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T20:50:11Z","title_canon_sha256":"7914667cb52bbe8399bf9b06c6f6602210d329337de07f9f49d84a12270c133f"},"schema_version":"1.0","source":{"id":"2502.19557","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.19557","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"arxiv_version","alias_value":"2502.19557v1","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.19557","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"pith_short_12","alias_value":"SRJEQMARHX3H","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"pith_short_16","alias_value":"SRJEQMARHX3HZREM","created_at":"2026-07-05T10:20:48Z"},{"alias_kind":"pith_short_8","alias_value":"SRJEQMAR","created_at":"2026-07-05T10:20:48Z"}],"graph_snapshots":[{"event_id":"sha256:ffd01babf75cb78f58e22fb307d7796e7ea16055fe4a10829e5152e7a789eaa1","target":"graph","created_at":"2026-07-05T10:20:48Z","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/2502.19557/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Distilling large language models (LLMs) typically involves transferring the teacher model's responses through supervised fine-tuning (SFT). However, this approach neglects the potential to distill both data (output content) and reward signals (quality evaluations). Extracting reliable reward signals directly from teacher models is challenging, as LLMs are optimized for generation rather than evaluation, often resulting in biased or inconsistent assessments. To address this limitation, we propose a novel distillation pipeline that transfers both responses and rewards. Our method generates pseud","authors_text":"Chenghua Huang, Dongmei Zhang, Jun Wang, Lu Wang, Meng Fang, Mykola Pechenizkiy, Qingwei Lin, Qi Zhang, Saravan Rajmohan, Yali Du, Yudi Zhang","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T20:50:11Z","title":"Distill Not Only Data but Also Rewards: Can Smaller Language Models Surpass Larger Ones?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.19557","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":"sha256:7c473f55ffcec25b4c9a7196feb184ad8fddb89dd73f2973cc2e8fc1bada983d","target":"record","created_at":"2026-07-05T10:20:48Z","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":"eb8eb4bac67a4a338eb4a7da4c7a77aadd8a5a5fb0cdc03d0ba28b68d98e6fc3","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T20:50:11Z","title_canon_sha256":"7914667cb52bbe8399bf9b06c6f6602210d329337de07f9f49d84a12270c133f"},"schema_version":"1.0","source":{"id":"2502.19557","kind":"arxiv","version":1}},"canonical_sha256":"94524830113df67cc48c8fb62d880fc6fdcfd90b404ef5adbef4f745724095e9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"94524830113df67cc48c8fb62d880fc6fdcfd90b404ef5adbef4f745724095e9","first_computed_at":"2026-07-05T10:20:48.674387Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:20:48.674387Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RibY8iONNyUwKIC0SNLZFMTEpW0S0qWVJARZa6H7Nwdvwdo3QfOx6RrYSRP3X+fZSFo6A9u/5W67+YHiMFOWCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:20:48.674896Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.19557","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7c473f55ffcec25b4c9a7196feb184ad8fddb89dd73f2973cc2e8fc1bada983d","sha256:ffd01babf75cb78f58e22fb307d7796e7ea16055fe4a10829e5152e7a789eaa1"],"state_sha256":"8060359633d9510641938a272a5de42ba0580e80bdc85620c9c241d0fba0f3c2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FgFbkaBDO3C7Z8Zxgq9+h6btUxmbwbd/T8N/3IygXldtqzn/drqfFJ7H2QyDou5dNtJB328yvKtPak2U8irkBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T11:38:57.612651Z","bundle_sha256":"bbb55932100cefdc8683d3107e77fccd82c4abf966e41313ebc3e8b9b0f64667"}}