{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GED7I23AFEIXJTSWRDZ4MPTSU7","short_pith_number":"pith:GED7I23A","schema_version":"1.0","canonical_sha256":"3107f46b60291174ce5688f3c63e72a7f848984183ddfdf8acb9e8c377e4f596","source":{"kind":"arxiv","id":"2607.17299","version":1},"attestation_state":"computed","paper":{"title":"WAR: Workload-Aware Rollouts for Synchronous Agentic Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.OS"],"primary_cat":"cs.LG","authors_text":"Atlas Zhao, David Bao, Frank Du, Ryan Xu","submitted_at":"2026-07-19T15:31:13Z","abstract_excerpt":"Long-horizon rollout generation has become the dominant systems bottleneck in agentic reinforcement learning (RL). As agents interact with environments over many turns, trajectories rapidly grow to tens of thousands of tokens, making synchronous RL training increasingly constrained by rollout. We propose WAR, a workload-aware rollout system that substantially accelerates synchronous agentic RL by jointly optimizing decoding and scheduling. WAR is built on a key observation: the optimal rollout optimization strategy depends on runtime load: (1) Under low load, WAR enables model-free speculative"},"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":"2607.17299","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-07-19T15:31:13Z","cross_cats_sorted":["cs.AI","cs.CL","cs.OS"],"title_canon_sha256":"6179aea31c47d07e0a8ba0a4cc6474ca3809a251f3d2bd4a9769bb8ecb39e483","abstract_canon_sha256":"7a7b7f52f76dec6e93dc49aaefe048ec642188ba1c1da607bf6a52a0d858223f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-21T01:21:26.056520Z","signature_b64":"Ydw1GfewtXhFUmemk0j30H0cligWIKWmCsnhOTkJlLHKuU8nEHcULxO3azFG0Z7tqR1kVJ9cQZ6HZiPqmOGiAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3107f46b60291174ce5688f3c63e72a7f848984183ddfdf8acb9e8c377e4f596","last_reissued_at":"2026-07-21T01:21:26.055679Z","signature_status":"signed_v1","first_computed_at":"2026-07-21T01:21:26.055679Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WAR: Workload-Aware Rollouts for Synchronous Agentic Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.OS"],"primary_cat":"cs.LG","authors_text":"Atlas Zhao, David Bao, Frank Du, Ryan Xu","submitted_at":"2026-07-19T15:31:13Z","abstract_excerpt":"Long-horizon rollout generation has become the dominant systems bottleneck in agentic reinforcement learning (RL). As agents interact with environments over many turns, trajectories rapidly grow to tens of thousands of tokens, making synchronous RL training increasingly constrained by rollout. We propose WAR, a workload-aware rollout system that substantially accelerates synchronous agentic RL by jointly optimizing decoding and scheduling. WAR is built on a key observation: the optimal rollout optimization strategy depends on runtime load: (1) Under low load, WAR enables model-free speculative"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.17299","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/2607.17299/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":"2607.17299","created_at":"2026-07-21T01:21:26.056110+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.17299v1","created_at":"2026-07-21T01:21:26.056110+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.17299","created_at":"2026-07-21T01:21:26.056110+00:00"},{"alias_kind":"pith_short_12","alias_value":"GED7I23AFEIX","created_at":"2026-07-21T01:21:26.056110+00:00"},{"alias_kind":"pith_short_16","alias_value":"GED7I23AFEIXJTSW","created_at":"2026-07-21T01:21:26.056110+00:00"},{"alias_kind":"pith_short_8","alias_value":"GED7I23A","created_at":"2026-07-21T01:21:26.056110+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/GED7I23AFEIXJTSWRDZ4MPTSU7","json":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7.json","graph_json":"https://pith.science/api/pith-number/GED7I23AFEIXJTSWRDZ4MPTSU7/graph.json","events_json":"https://pith.science/api/pith-number/GED7I23AFEIXJTSWRDZ4MPTSU7/events.json","paper":"https://pith.science/paper/GED7I23A"},"agent_actions":{"view_html":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7","download_json":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7.json","view_paper":"https://pith.science/paper/GED7I23A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.17299&json=true","fetch_graph":"https://pith.science/api/pith-number/GED7I23AFEIXJTSWRDZ4MPTSU7/graph.json","fetch_events":"https://pith.science/api/pith-number/GED7I23AFEIXJTSWRDZ4MPTSU7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7/action/storage_attestation","attest_author":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7/action/author_attestation","sign_citation":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7/action/citation_signature","submit_replication":"https://pith.science/pith/GED7I23AFEIXJTSWRDZ4MPTSU7/action/replication_record"}},"created_at":"2026-07-21T01:21:26.056110+00:00","updated_at":"2026-07-21T01:21:26.056110+00:00"}