{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:HJUBM5TT7WVWLRABEHAP6CO5PC","short_pith_number":"pith:HJUBM5TT","schema_version":"1.0","canonical_sha256":"3a68167673fdab65c40121c0ff09dd78adaa5a1bd0727c9004b7f050811da989","source":{"kind":"arxiv","id":"2602.05472","version":2},"attestation_state":"computed","paper":{"title":"ALIVE: Awakening LLM Reasoning via Adversarial Learning and Instructive Verbal Evaluation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jing Ye, Xinpei Zhao, Yiwen Duan","submitted_at":"2026-02-05T09:20:23Z","abstract_excerpt":"The quest for expert-level reasoning in Large Language Models (LLMs) has been hampered by a persistent \\textit{reward bottleneck}: traditional reinforcement learning (RL) relies on scalar rewards that are \\textbf{costly} to scale, \\textbf{brittle} across domains, and \\textbf{blind} to the underlying logic of a solution. This reliance on external, impoverished signals prevents models from developing a deep, self-contained understanding of reasoning principles. We introduce \\textbf{ALIVE} (\\emph{Adversarial Learning with Instructive Verbal Evaluation}), a hands-free alignment framework that move"},"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":"2602.05472","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-02-05T09:20:23Z","cross_cats_sorted":[],"title_canon_sha256":"995dd70a6e68d89f98112bae745d08422bcf80ecb4a52e215c12c40db9a2e0fe","abstract_canon_sha256":"365eb6d80e6a0c6b85d2a1e052510e7c0ada13d53de57823308fc8c156561cdf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:14.447276Z","signature_b64":"uiugWSSb0MVBqhH9Mz1S+ADJKBjXdKB76WYULwhRLtuuN51rwr7+6MlgTuCOJG+tVKrV+TF9sp7vQT4cmcpbCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3a68167673fdab65c40121c0ff09dd78adaa5a1bd0727c9004b7f050811da989","last_reissued_at":"2026-05-25T02:01:14.446293Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:14.446293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ALIVE: Awakening LLM Reasoning via Adversarial Learning and Instructive Verbal Evaluation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jing Ye, Xinpei Zhao, Yiwen Duan","submitted_at":"2026-02-05T09:20:23Z","abstract_excerpt":"The quest for expert-level reasoning in Large Language Models (LLMs) has been hampered by a persistent \\textit{reward bottleneck}: traditional reinforcement learning (RL) relies on scalar rewards that are \\textbf{costly} to scale, \\textbf{brittle} across domains, and \\textbf{blind} to the underlying logic of a solution. This reliance on external, impoverished signals prevents models from developing a deep, self-contained understanding of reasoning principles. We introduce \\textbf{ALIVE} (\\emph{Adversarial Learning with Instructive Verbal Evaluation}), a hands-free alignment framework that move"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.05472","kind":"arxiv","version":2},"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/2602.05472/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":"2602.05472","created_at":"2026-05-25T02:01:14.446414+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.05472v2","created_at":"2026-05-25T02:01:14.446414+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.05472","created_at":"2026-05-25T02:01:14.446414+00:00"},{"alias_kind":"pith_short_12","alias_value":"HJUBM5TT7WVW","created_at":"2026-05-25T02:01:14.446414+00:00"},{"alias_kind":"pith_short_16","alias_value":"HJUBM5TT7WVWLRAB","created_at":"2026-05-25T02:01:14.446414+00:00"},{"alias_kind":"pith_short_8","alias_value":"HJUBM5TT","created_at":"2026-05-25T02:01:14.446414+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.16727","citing_title":"PopuLoRA: Co-Evolving LLM Populations for Reasoning Self-Play","ref_index":44,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC","json":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC.json","graph_json":"https://pith.science/api/pith-number/HJUBM5TT7WVWLRABEHAP6CO5PC/graph.json","events_json":"https://pith.science/api/pith-number/HJUBM5TT7WVWLRABEHAP6CO5PC/events.json","paper":"https://pith.science/paper/HJUBM5TT"},"agent_actions":{"view_html":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC","download_json":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC.json","view_paper":"https://pith.science/paper/HJUBM5TT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.05472&json=true","fetch_graph":"https://pith.science/api/pith-number/HJUBM5TT7WVWLRABEHAP6CO5PC/graph.json","fetch_events":"https://pith.science/api/pith-number/HJUBM5TT7WVWLRABEHAP6CO5PC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC/action/storage_attestation","attest_author":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC/action/author_attestation","sign_citation":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC/action/citation_signature","submit_replication":"https://pith.science/pith/HJUBM5TT7WVWLRABEHAP6CO5PC/action/replication_record"}},"created_at":"2026-05-25T02:01:14.446414+00:00","updated_at":"2026-05-25T02:01:14.446414+00:00"}