{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VSDCZAFQEYNX7AUB3ZURGFKPPS","short_pith_number":"pith:VSDCZAFQ","schema_version":"1.0","canonical_sha256":"ac862c80b0261b7f8281de6913154f7ca92cf8a19448f6bcfec847830c704b2a","source":{"kind":"arxiv","id":"2605.22537","version":1},"attestation_state":"computed","paper":{"title":"F-TIS: Harnessing Diverse Models in Collaborative GRPO","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lydia Yiyu Chen, Nikolay Blagoev, O\\u{g}uzhan Ersoy, Wendelin Boehmer","submitted_at":"2026-05-21T14:25:24Z","abstract_excerpt":"Reinforcement learning methods such as GRPO have seen great popularity in LLM post-training. In GRPO, models produce completions to a set of prompts, which are rewarded, and the policy is updated towards the relatively high reward completions. Due to the auto-regressive nature of models, the generation phase of such style of training can be extremely time consuming. As a solution, prior work has sought to distribute the inference step across many nodes, working parallel. These works assume primarily homogeneous models in the training in order to keep samples as close to on-policy as possible. "},"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":"2605.22537","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-21T14:25:24Z","cross_cats_sorted":[],"title_canon_sha256":"4b46c646abf025917587a74a780a70fb1544a80562eed8d24b857923c1aed126","abstract_canon_sha256":"5b2da5db38aaa50ca279d8791cc1fa2bde21ec1f61a6c7d4d4e54ab7de603b47"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:04:56.250851Z","signature_b64":"f34w9UUtMzkvGp3bn3fOLWdb8K1zc4JhTvuZAVI8UZvqqtru39sI3XlIMY99hmjPsMFMC0VybQn05aXsvlE1Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ac862c80b0261b7f8281de6913154f7ca92cf8a19448f6bcfec847830c704b2a","last_reissued_at":"2026-05-22T01:04:56.250064Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:04:56.250064Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"F-TIS: Harnessing Diverse Models in Collaborative GRPO","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Lydia Yiyu Chen, Nikolay Blagoev, O\\u{g}uzhan Ersoy, Wendelin Boehmer","submitted_at":"2026-05-21T14:25:24Z","abstract_excerpt":"Reinforcement learning methods such as GRPO have seen great popularity in LLM post-training. In GRPO, models produce completions to a set of prompts, which are rewarded, and the policy is updated towards the relatively high reward completions. Due to the auto-regressive nature of models, the generation phase of such style of training can be extremely time consuming. As a solution, prior work has sought to distribute the inference step across many nodes, working parallel. These works assume primarily homogeneous models in the training in order to keep samples as close to on-policy as possible. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.22537","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/2605.22537/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":"2605.22537","created_at":"2026-05-22T01:04:56.250171+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.22537v1","created_at":"2026-05-22T01:04:56.250171+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.22537","created_at":"2026-05-22T01:04:56.250171+00:00"},{"alias_kind":"pith_short_12","alias_value":"VSDCZAFQEYNX","created_at":"2026-05-22T01:04:56.250171+00:00"},{"alias_kind":"pith_short_16","alias_value":"VSDCZAFQEYNX7AUB","created_at":"2026-05-22T01:04:56.250171+00:00"},{"alias_kind":"pith_short_8","alias_value":"VSDCZAFQ","created_at":"2026-05-22T01:04:56.250171+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/VSDCZAFQEYNX7AUB3ZURGFKPPS","json":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS.json","graph_json":"https://pith.science/api/pith-number/VSDCZAFQEYNX7AUB3ZURGFKPPS/graph.json","events_json":"https://pith.science/api/pith-number/VSDCZAFQEYNX7AUB3ZURGFKPPS/events.json","paper":"https://pith.science/paper/VSDCZAFQ"},"agent_actions":{"view_html":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS","download_json":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS.json","view_paper":"https://pith.science/paper/VSDCZAFQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.22537&json=true","fetch_graph":"https://pith.science/api/pith-number/VSDCZAFQEYNX7AUB3ZURGFKPPS/graph.json","fetch_events":"https://pith.science/api/pith-number/VSDCZAFQEYNX7AUB3ZURGFKPPS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS/action/storage_attestation","attest_author":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS/action/author_attestation","sign_citation":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS/action/citation_signature","submit_replication":"https://pith.science/pith/VSDCZAFQEYNX7AUB3ZURGFKPPS/action/replication_record"}},"created_at":"2026-05-22T01:04:56.250171+00:00","updated_at":"2026-05-22T01:04:56.250171+00:00"}