{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:IE22MEPKVMN7TVPLV7Q6ZZG5VE","short_pith_number":"pith:IE22MEPK","schema_version":"1.0","canonical_sha256":"4135a611eaab1bf9d5ebafe1ece4dda910479941188858f87674c26c19c27f37","source":{"kind":"arxiv","id":"2607.09207","version":1},"attestation_state":"computed","paper":{"title":"Bidirectional Resource Scheduling for Disaggregated and Asynchronous RL Post-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Chu Xiaowen, Shi Shaohuai, Tan Zhiqiang, Wang Maoxin, Wang Qiang, Wang Sijie, Yin Yiming","submitted_at":"2026-07-10T08:51:14Z","abstract_excerpt":"It is well established that the reasoning capabilities of large language models (LLMs) can be improved by applying reinforcement learning (RL) in a post-training stage. In a standard RL iteration, the current model (the policy) generates experience through rollouts, and the resulting data is then used to update the policy during training. High-performance RL frameworks such as StreamRL and AReaL employ a disaggregated architecture and asynchronous rollouts to better exploit both rollout and training resources, thereby increasing overall system throughput.\n  Nonetheless, across varying RL setup"},"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":"2607.09207","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-07-10T08:51:14Z","cross_cats_sorted":[],"title_canon_sha256":"b4d91c4b6a31ce6826b665aaa04220881cef2cf38859f200f1f02457077ebf95","abstract_canon_sha256":"9ae4d780e79d948d834998944d94e80b3f15626006a783127c3593f889d2c89a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-13T01:18:51.447406Z","signature_b64":"vTCvtXWHSM5Y+dCF4vUt2krHMSDRILmCys+CJhv92k+8VLvQ2lK47eOSAbluRPhr2z6JqvUdyWP9y7fesCWVAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4135a611eaab1bf9d5ebafe1ece4dda910479941188858f87674c26c19c27f37","last_reissued_at":"2026-07-13T01:18:51.445918Z","signature_status":"signed_v1","first_computed_at":"2026-07-13T01:18:51.445918Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bidirectional Resource Scheduling for Disaggregated and Asynchronous RL Post-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Chu Xiaowen, Shi Shaohuai, Tan Zhiqiang, Wang Maoxin, Wang Qiang, Wang Sijie, Yin Yiming","submitted_at":"2026-07-10T08:51:14Z","abstract_excerpt":"It is well established that the reasoning capabilities of large language models (LLMs) can be improved by applying reinforcement learning (RL) in a post-training stage. In a standard RL iteration, the current model (the policy) generates experience through rollouts, and the resulting data is then used to update the policy during training. High-performance RL frameworks such as StreamRL and AReaL employ a disaggregated architecture and asynchronous rollouts to better exploit both rollout and training resources, thereby increasing overall system throughput.\n  Nonetheless, across varying RL setup"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.09207","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/2607.09207/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":"2607.09207","created_at":"2026-07-13T01:18:51.446799+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.09207v1","created_at":"2026-07-13T01:18:51.446799+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.09207","created_at":"2026-07-13T01:18:51.446799+00:00"},{"alias_kind":"pith_short_12","alias_value":"IE22MEPKVMN7","created_at":"2026-07-13T01:18:51.446799+00:00"},{"alias_kind":"pith_short_16","alias_value":"IE22MEPKVMN7TVPL","created_at":"2026-07-13T01:18:51.446799+00:00"},{"alias_kind":"pith_short_8","alias_value":"IE22MEPK","created_at":"2026-07-13T01:18:51.446799+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/IE22MEPKVMN7TVPLV7Q6ZZG5VE","json":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE.json","graph_json":"https://pith.science/api/pith-number/IE22MEPKVMN7TVPLV7Q6ZZG5VE/graph.json","events_json":"https://pith.science/api/pith-number/IE22MEPKVMN7TVPLV7Q6ZZG5VE/events.json","paper":"https://pith.science/paper/IE22MEPK"},"agent_actions":{"view_html":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE","download_json":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE.json","view_paper":"https://pith.science/paper/IE22MEPK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.09207&json=true","fetch_graph":"https://pith.science/api/pith-number/IE22MEPKVMN7TVPLV7Q6ZZG5VE/graph.json","fetch_events":"https://pith.science/api/pith-number/IE22MEPKVMN7TVPLV7Q6ZZG5VE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE/action/storage_attestation","attest_author":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE/action/author_attestation","sign_citation":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE/action/citation_signature","submit_replication":"https://pith.science/pith/IE22MEPKVMN7TVPLV7Q6ZZG5VE/action/replication_record"}},"created_at":"2026-07-13T01:18:51.446799+00:00","updated_at":"2026-07-13T01:18:51.446799+00:00"}