{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:XST5JYQVSI4FPDLYXGWK54TIZM","short_pith_number":"pith:XST5JYQV","schema_version":"1.0","canonical_sha256":"bca7d4e2159238578d78b9acaef268cb0e75df3e4af423f0efc36263586f9382","source":{"kind":"arxiv","id":"2606.12507","version":1},"attestation_state":"computed","paper":{"title":"Rubric-Guided Self-Distillation: Post-Training Without Rubric Verifiers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aakash Sabharwal, Advait Gosai, Anas Mahmoud, Bing Liu, MohammadHossein Rezaei, Razvan-Gabriel Dumitru, Utkarsh Tyagi, Yunzhong He, Zihao Wang","submitted_at":"2026-06-10T17:53:19Z","abstract_excerpt":"Rubrics have emerged as an alternative to RLVR in open-ended domains where a single ground-truth final answer is not available. Existing rubric-based training methods rely on an LLM verifier that scores each rollout against rubrics. This introduces substantial training-time overhead, exposes optimization to verifier-specific biases, and reduces rubric feedback to a sparse end-of-trajectory signal. We propose Rubric-Guided Self-Distillation (RGSD), a verifier-free training method in which the base policy, conditioned on the rubric, serves as the teacher for the unconditioned student. RGSD disti"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.12507","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-10T17:53:19Z","cross_cats_sorted":[],"title_canon_sha256":"b44b4bff8a37f604baa7e3c3d717995ff2c96f56618893e4c78ab93912169728","abstract_canon_sha256":"c01b6e96176d37d5b3badd0383b386456a1845e089f41653ac6a98c63b6a525e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T00:07:52.785569Z","signature_b64":"g+iEmUzkFFm9oyHTiX5Ary7xaMWUKWrmRapLGgy4spBCEsZLJTSCfyUIFwLqinXl2yajUJG0KzrCLX/zVRyCBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bca7d4e2159238578d78b9acaef268cb0e75df3e4af423f0efc36263586f9382","last_reissued_at":"2026-06-12T00:07:52.785145Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T00:07:52.785145Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rubric-Guided Self-Distillation: Post-Training Without Rubric Verifiers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aakash Sabharwal, Advait Gosai, Anas Mahmoud, Bing Liu, MohammadHossein Rezaei, Razvan-Gabriel Dumitru, Utkarsh Tyagi, Yunzhong He, Zihao Wang","submitted_at":"2026-06-10T17:53:19Z","abstract_excerpt":"Rubrics have emerged as an alternative to RLVR in open-ended domains where a single ground-truth final answer is not available. Existing rubric-based training methods rely on an LLM verifier that scores each rollout against rubrics. This introduces substantial training-time overhead, exposes optimization to verifier-specific biases, and reduces rubric feedback to a sparse end-of-trajectory signal. We propose Rubric-Guided Self-Distillation (RGSD), a verifier-free training method in which the base policy, conditioned on the rubric, serves as the teacher for the unconditioned student. RGSD disti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.12507","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/2606.12507/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.12507","created_at":"2026-06-12T00:07:52.785202+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.12507v1","created_at":"2026-06-12T00:07:52.785202+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.12507","created_at":"2026-06-12T00:07:52.785202+00:00"},{"alias_kind":"pith_short_12","alias_value":"XST5JYQVSI4F","created_at":"2026-06-12T00:07:52.785202+00:00"},{"alias_kind":"pith_short_16","alias_value":"XST5JYQVSI4FPDLY","created_at":"2026-06-12T00:07:52.785202+00:00"},{"alias_kind":"pith_short_8","alias_value":"XST5JYQV","created_at":"2026-06-12T00:07:52.785202+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM","json":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM.json","graph_json":"https://pith.science/api/pith-number/XST5JYQVSI4FPDLYXGWK54TIZM/graph.json","events_json":"https://pith.science/api/pith-number/XST5JYQVSI4FPDLYXGWK54TIZM/events.json","paper":"https://pith.science/paper/XST5JYQV"},"agent_actions":{"view_html":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM","download_json":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM.json","view_paper":"https://pith.science/paper/XST5JYQV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.12507&json=true","fetch_graph":"https://pith.science/api/pith-number/XST5JYQVSI4FPDLYXGWK54TIZM/graph.json","fetch_events":"https://pith.science/api/pith-number/XST5JYQVSI4FPDLYXGWK54TIZM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM/action/storage_attestation","attest_author":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM/action/author_attestation","sign_citation":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM/action/citation_signature","submit_replication":"https://pith.science/pith/XST5JYQVSI4FPDLYXGWK54TIZM/action/replication_record"}},"created_at":"2026-06-12T00:07:52.785202+00:00","updated_at":"2026-06-12T00:07:52.785202+00:00"}