{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:K7PQLHJXOLT42AEJT4XOXPFDVW","short_pith_number":"pith:K7PQLHJX","schema_version":"1.0","canonical_sha256":"57df059d3772e7cd00899f2eebbca3ada7cc154f164b5feb35136fac7523a830","source":{"kind":"arxiv","id":"2608.08224","version":1},"attestation_state":"computed","paper":{"title":"Control-Diverse Reinforcement Fine-Tuning: Decoupling the Shared Control Bottleneck of RL Post-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Atsushi Koike, Binwen Tan, Dengzhe Hou, Fangzhou Lin, Jingchao Wang, Kazunori Yamada, Lingyu Jiang, Yunhan Shen, Zeyuan Wu","submitted_at":"2026-08-08T16:44:47Z","abstract_excerpt":"Reinforcement learning post-training unlocks complex reasoning in LLMs. Yet benchmark scores reveal only whether a model improved, not what changed inside it, nor how it splits finite capability across tasks. A representative interpretability line attributes the success of RL fine-tuning to stronger and more diverse circuit activation. We challenge this activation-centered account by separating activation from control: an activated circuit need not control the post-training reward gain. Adapting Metabolic Control Analysis, we define the Post-training Control Coefficient to measure component co"},"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":"2608.08224","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-08-08T16:44:47Z","cross_cats_sorted":[],"title_canon_sha256":"103de1a764a7df596fc5f468b2a455efa96c4ea4fe837f7e58209c32cc2d46e9","abstract_canon_sha256":"40827ced7a91d8c1cb1a46329e6c14aa038604b6ccb702bb0c38cdf94c0395f6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:22:03.503755Z","signature_b64":"S3eb01PX2wS3m9rIg9V5xhzzRgVfdeZLLagS2O0peWJ8sJH7Y7ZEymJBfGBlqP+zBGmDhUqx+1B1jBGYjub1CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"57df059d3772e7cd00899f2eebbca3ada7cc154f164b5feb35136fac7523a830","last_reissued_at":"2026-08-11T01:22:03.501352Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:22:03.501352Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Control-Diverse Reinforcement Fine-Tuning: Decoupling the Shared Control Bottleneck of RL Post-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Atsushi Koike, Binwen Tan, Dengzhe Hou, Fangzhou Lin, Jingchao Wang, Kazunori Yamada, Lingyu Jiang, Yunhan Shen, Zeyuan Wu","submitted_at":"2026-08-08T16:44:47Z","abstract_excerpt":"Reinforcement learning post-training unlocks complex reasoning in LLMs. Yet benchmark scores reveal only whether a model improved, not what changed inside it, nor how it splits finite capability across tasks. A representative interpretability line attributes the success of RL fine-tuning to stronger and more diverse circuit activation. We challenge this activation-centered account by separating activation from control: an activated circuit need not control the post-training reward gain. Adapting Metabolic Control Analysis, we define the Post-training Control Coefficient to measure component co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.08224","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/2608.08224/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":"2608.08224","created_at":"2026-08-11T01:22:03.502145+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.08224v1","created_at":"2026-08-11T01:22:03.502145+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.08224","created_at":"2026-08-11T01:22:03.502145+00:00"},{"alias_kind":"pith_short_12","alias_value":"K7PQLHJXOLT4","created_at":"2026-08-11T01:22:03.502145+00:00"},{"alias_kind":"pith_short_16","alias_value":"K7PQLHJXOLT42AEJ","created_at":"2026-08-11T01:22:03.502145+00:00"},{"alias_kind":"pith_short_8","alias_value":"K7PQLHJX","created_at":"2026-08-11T01:22:03.502145+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/K7PQLHJXOLT42AEJT4XOXPFDVW","json":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW.json","graph_json":"https://pith.science/api/pith-number/K7PQLHJXOLT42AEJT4XOXPFDVW/graph.json","events_json":"https://pith.science/api/pith-number/K7PQLHJXOLT42AEJT4XOXPFDVW/events.json","paper":"https://pith.science/paper/K7PQLHJX"},"agent_actions":{"view_html":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW","download_json":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW.json","view_paper":"https://pith.science/paper/K7PQLHJX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.08224&json=true","fetch_graph":"https://pith.science/api/pith-number/K7PQLHJXOLT42AEJT4XOXPFDVW/graph.json","fetch_events":"https://pith.science/api/pith-number/K7PQLHJXOLT42AEJT4XOXPFDVW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW/action/storage_attestation","attest_author":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW/action/author_attestation","sign_citation":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW/action/citation_signature","submit_replication":"https://pith.science/pith/K7PQLHJXOLT42AEJT4XOXPFDVW/action/replication_record"}},"created_at":"2026-08-11T01:22:03.502145+00:00","updated_at":"2026-08-11T01:22:03.502145+00:00"}