{"_canonicalization":{"envelope_id":"axm_ + sha256(envelope minus {signature, axiom_id, anchors})","envelope_signature":"ed25519(envelope minus {signature, axiom_id})","json":"sort_keys=True, separators=(',',':'), ensure_ascii=False, allow_nan=False, utf-8","leaf_hash":"sha256(0x00 || canonical_json(envelope_full))","seal_signature":"ed25519(seal minus {signature, sig_algorithm})"},"axiom_id":"axm_f3ce631040de2ee5c665f717f98f20341a5d11bfc1de1761f693d4056c3a0c09","bitcoin_anchor":{"bitcoin_attestations":[],"calendar_attestations":[],"ots_url":"","stamped_at":"","status":"pending_next_stamp"},"envelope":{"anchors":[{"chain":"crovia.axiom_graph","height":0,"merkle_proof":"spider_vendor_press_v1","root_at_anchor":"spider_vendor_press_v1"}],"axiom_id":"axm_f3ce631040de2ee5c665f717f98f20341a5d11bfc1de1761f693d4056c3a0c09","axiom_type":"AX.OBS","body":{"axiom_subtype":"news.vendor_press.v1","category":"news","fingerprint":"e77bd569d81d77d13f5f346d883315277e0d664e263bc1e42ece4ced97e588b8","published":"Wed, 10 Jun 2026 00:00:00 -0400","receipt_hash":"e77bd569d81d77d13f5f346d883315277e0d664e263bc1e42ece4ced97e588b8","schema":"spider.news.vendor_press.v1","spider":"vendor_press","spider_record":{"axiom_subtype":"news.vendor_press.v1","category":"news","decision_hint":"POSITIVE","envelope_target":"AX.OBS","fingerprint":"e77bd569d81d77d13f5f346d883315277e0d664e263bc1e42ece4ced97e588b8","observed_at":"2026-06-10T04:43:37.461885Z","parent_run_hash":"23aff1a6f676ba7ca33f70f4ddfae1dd282fb86104d577ce9be510d81a94c5dc","published":"Wed, 10 Jun 2026 00:00:00 -0400","runtime_version":"0.1.0","schema":"spider.news.vendor_press.v1","source_status":200,"source_url":"https://export.arxiv.org/rss/cs.AI","spider":"vendor_press","summary_excerpt":"arXiv:2606.09932v1 Announce Type: cross \nAbstract: Supervised Fine-Tuning (SFT) followed by Reinforcement Learning (RL) has become a standard pipeline for Large Language Model (LLM) post-training. SFT is expected to provide a useful behavioral prior for RL to further enhance model capabilities. However, checkpoints with excessive SFT often show limited improvement during RL. We attribute this failure to the loss of model plasticity: the reduced ability of an SFT-initialized policy to be effectively reshaped by subsequent RL. To better understand this phenomenon, we conduct detailed analysis from multiple perspectives, including parameter changes, output spaces, and RL optimization dynamics. Our results show that models from excessive SFT tend to produce over-confident token distributions and exhibit sharp parameter landscapes, which make them harder to optimize in the RL stage. To enable a more robust SFT-to-RL handoff, we propose \\texttt{Rejuvenation}, a simple yet effective method th","title":"When RL Fails after SFT: Rejuvenating Model Plasticity for Robust SFT-to-RL Handoff","url":"https://arxiv.org/abs/2606.09932","vendor":"arxiv_cs_ai"},"summary":"arXiv:2606.09932v1 Announce Type: cross \nAbstract: Supervised Fine-Tuning (SFT) followed by Reinforcement Learning (RL) has become a standard pipeline for Large Language Model (LLM) post-training. SFT is expected to provide a useful behavioral prior for RL to further enhance model capabilities. However, checkpoints with excessive SFT often show limited improvement during RL. We attribute this failure to the loss of model plasticity: the reduced ability of an SFT-initialized policy to be effectively reshaped by subsequent RL. To better understand this phenomenon, we conduct detailed analysis from multiple perspectives, including parameter changes, output spaces, and RL optimization dynamics. Our results show that models from excessive SFT tend to produce over-confident token distributions and exhibit sharp parameter landscapes, which make them harder to optimize in the RL stage. To enable a more robust SFT-to-RL handoff, we propose \\texttt{Rejuvenation}, a simple yet effective method th","title":"When RL Fails after SFT: Rejuvenating Model Plasticity for Robust SFT-to-RL Handoff","vendor":"arxiv_cs_ai"},"confidence":{"method":"deterministic"},"decision":"POSITIVE","issued_at":"2026-06-10T04:43:37Z","notes":"Spider vendor_press (news) news.vendor_press.v1","object":{"captured_by":"crovia.spider.vendor_press","primary_source_url":"https://arxiv.org/abs/2606.09932"},"predecessors":[],"schema":"crovia.axiom.v1","signature":"ed25519:78fb45458d577e8a10407838f17ecd8d89e1160f479eb7ee024610d4d6db849472b34b7dfa6b346c172158ffab020e2ef599ec1965beffbfc4c194e10646b608","signer":"crovia.substrate","subject":{"observed_at":"2026-06-10T04:43:37Z","source_collector":"spider:vendor_press","target_id":"https://arxiv.org/abs/2606.09932"},"tsa":{"authority":"crovia.substrate.bootstrap","rfc3161_token":"{\"kind\":\"crovia.bootstrap.tsa\",\"source_jsonl\":\"/opt/crovia/spider/data/news/vendor_press_v1.jsonl\",\"source_seal_merkle_root\":\"spider_vendor_press_v1\",\"upgrade_path\":\"Sessione H \\u2014 OpenTimestamps weekly anchor\"}"},"zk_mode":"clear","zk_proof":null},"ledger":{"leaf_hash":"59f24ad6b802f0c1bd7ac68d108a2bc0d244dd9c4b6d66e96e02d38789aeb5b7","leaf_index":225951,"ledger_path":"/opt/crovia/substrate/axiom_ledger.jsonl"},"merkle_proof":{"hash_alg":"sha256","leaf_prefix":"0x00","node_prefix":"0x01","odd_leaf_rule":"duplicate_last","path":[{"sibling":"f212a7a23a81f5b1ae3049dcc96331e8715839f22ad30387f768a9f6dfa285c2","side":"left"},{"sibling":"b2313ec31b767394765cf606af3db35dba532b35591d21dbfeefbcf1e6f95bf0","side":"left"},{"sibling":"ffba732a8bb29ba74de6a8b89e01690a85e0ddcd64c2e2b0b2379163de9d08d3","side":"left"},{"sibling":"511f5462f15a165f70b99cca0ea8b3ab413ca1b19cf0221b0c66e044df53643f","side":"left"},{"sibling":"4fd2b5823301dc15427ce56aaf8b1445ec64389598421ab6286e94add0e4ced7","side":"left"},{"sibling":"008c026cfe61c71a5051aee5c9d0fe3db37fa3a8c80d637546b0563137be0b3f","side":"right"},{"sibling":"4189fb616853b63efeccd8c98308b970a4444cb48cec9d4d57d47996103bbc76","side":"right"},{"sibling":"3d9f311f467db31ee1be2d9664daf3ef6547a3814a86217d383241d0655ab769","side":"left"},{"sibling":"ac698c3a6027f35b513fe892f166e70344e855625238cbb581d0bf06c131db80","side":"right"},{"sibling":"a4d17aefe58175050dc159af6246658fcf1c9f3ed57aacf1b350fc3261de4e69","side":"left"},{"sibling":"280b980aa0c7756b0b0cb22658f26466d36f0e70fbc3312cd2311d9898e30b8f","side":"right"},{"sibling":"c98954d4b658b1dda60fe52576fcf9bf21a2d49c67fb63f8c30f16ab5f721938","side":"right"},{"sibling":"cdb58f86163046d3b15f857b03372ec75e1ad9ea4548e086793d528b9eed364d","side":"left"},{"sibling":"533d82482604463aa4a281b9d8b85917b383b7c5f924b2b494039524c55e8797","side":"left"},{"sibling":"b2590791b920ca2a4ed39de126d2c0b1a10d9e7e62f572f12425f214e767b6e1","side":"left"},{"sibling":"6cea4964f32722eb370847c2f7c9d6a9f0622c239538b07e6815a59d6fd8d49c","side":"right"},{"sibling":"c300cf0154c136afc09b1702a0be98f4ba5b6dc5cf57e8cc714ec1eaf4196eff","side":"left"},{"sibling":"d841ad93efda0869e5eb97678f348f03f5caab4353e05ff4bf18f47fb945b822","side":"left"}]},"schema":"crovia.axiom_proof.v1","seal":{"first_collector_run_id":"","first_receipt_hash":"","jsonl_path":"/opt/crovia/substrate/axiom_ledger.jsonl","key_id":"430895f101d38164","last_collector_run_id":"","last_receipt_hash":"","leaf_count":228173,"merkle_root":"7e416202c0bfd759bd2eea4236713b403993d99793fe8badb5065040080bece3","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","run_id":"hourly_json_retrofit_20260611T143708Z","schema":"crovia.seal.v1","seal_family_version":"crovia-seal-family/1","seal_kind":"substrate_batch","sealed_at":"2026-06-11T21:59:35Z","sig_algorithm":"ed25519","signature":"231c80024bc3982dd493c45b31af95097e97aabc6d712a4e5bad7d0cbdd3c08e01ff395b0f8e72754bac97016e0cd0eed88b8a13cb71edbbcb9b6d72c10a7b03","signer_version":"1.1.0"},"trust_root":{"key_id":"430895f101d38164","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","signature_algorithm":"ed25519","url":"/registry/canon/TRUST_ROOT.md"},"verifier":{"spec":"/registry/canon/AXIOM_RECEIPT_v1.md","url":"/v/axm_f3ce631040de2ee5c665f717f98f20341a5d11bfc1de1761f693d4056c3a0c09"}}