{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:O26AE6SYFWDU2FMK7QLGCZA36G","short_pith_number":"pith:O26AE6SY","schema_version":"1.0","canonical_sha256":"76bc027a582d874d158afc1661641bf19bdaccd5152d41a2e11a6d269070be23","source":{"kind":"arxiv","id":"2606.31575","version":1},"attestation_state":"computed","paper":{"title":"Which Tokens Matter? Adaptive Token Selection for RLVR with the Relative Surprisal Index","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Baohua Dong, Hangcheng Zhu, Outongyi Lv, Xingjun Wang, Yanzhao Zheng, Yingda Chen, Yuanwei Zhang, Zhenghao Huang","submitted_at":"2026-06-30T12:33:19Z","abstract_excerpt":"Reinforcement learning (RL) has become a powerful tool for propelling Large Language Models (LLMs) beyond imitation-based training towards more robust reasoning capabilities. Among existing approaches, RL with Verifiable Rewards (RLVR) has emerged as a pivotal paradigm for advancing LLM reasoning. Despite its empirical success, recent studies have offered different insights. One line of inquiry advocates prioritizing high-entropy token positions during training, while another perspective cautions against allowing low-probability tokens to dominate gradient updates. Notably, although high-entro"},"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":"2606.31575","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-30T12:33:19Z","cross_cats_sorted":[],"title_canon_sha256":"77082c0ebe7a6ab8388ec03ec0e0bcf9863e4af0c1cd009a28f4ace8fa4c10c9","abstract_canon_sha256":"26335271c303d38e8b53dc28a980516ca1fee40493b3ded5f8234174f589bb0b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-01T01:18:08.128651Z","signature_b64":"z8B7P+huykce7s9jCl2XXSvZ6Pg80z/z0xJ37IYrOvUiC+UhNRaC0wKcXz5cpapxhTIUUH4rWerAzFk6oN2cCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"76bc027a582d874d158afc1661641bf19bdaccd5152d41a2e11a6d269070be23","last_reissued_at":"2026-07-01T01:18:08.128243Z","signature_status":"signed_v1","first_computed_at":"2026-07-01T01:18:08.128243Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Which Tokens Matter? Adaptive Token Selection for RLVR with the Relative Surprisal Index","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Baohua Dong, Hangcheng Zhu, Outongyi Lv, Xingjun Wang, Yanzhao Zheng, Yingda Chen, Yuanwei Zhang, Zhenghao Huang","submitted_at":"2026-06-30T12:33:19Z","abstract_excerpt":"Reinforcement learning (RL) has become a powerful tool for propelling Large Language Models (LLMs) beyond imitation-based training towards more robust reasoning capabilities. Among existing approaches, RL with Verifiable Rewards (RLVR) has emerged as a pivotal paradigm for advancing LLM reasoning. Despite its empirical success, recent studies have offered different insights. One line of inquiry advocates prioritizing high-entropy token positions during training, while another perspective cautions against allowing low-probability tokens to dominate gradient updates. Notably, although high-entro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.31575","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/2606.31575/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":"2606.31575","created_at":"2026-07-01T01:18:08.128298+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.31575v1","created_at":"2026-07-01T01:18:08.128298+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.31575","created_at":"2026-07-01T01:18:08.128298+00:00"},{"alias_kind":"pith_short_12","alias_value":"O26AE6SYFWDU","created_at":"2026-07-01T01:18:08.128298+00:00"},{"alias_kind":"pith_short_16","alias_value":"O26AE6SYFWDU2FMK","created_at":"2026-07-01T01:18:08.128298+00:00"},{"alias_kind":"pith_short_8","alias_value":"O26AE6SY","created_at":"2026-07-01T01:18:08.128298+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/O26AE6SYFWDU2FMK7QLGCZA36G","json":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G.json","graph_json":"https://pith.science/api/pith-number/O26AE6SYFWDU2FMK7QLGCZA36G/graph.json","events_json":"https://pith.science/api/pith-number/O26AE6SYFWDU2FMK7QLGCZA36G/events.json","paper":"https://pith.science/paper/O26AE6SY"},"agent_actions":{"view_html":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G","download_json":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G.json","view_paper":"https://pith.science/paper/O26AE6SY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.31575&json=true","fetch_graph":"https://pith.science/api/pith-number/O26AE6SYFWDU2FMK7QLGCZA36G/graph.json","fetch_events":"https://pith.science/api/pith-number/O26AE6SYFWDU2FMK7QLGCZA36G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G/action/storage_attestation","attest_author":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G/action/author_attestation","sign_citation":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G/action/citation_signature","submit_replication":"https://pith.science/pith/O26AE6SYFWDU2FMK7QLGCZA36G/action/replication_record"}},"created_at":"2026-07-01T01:18:08.128298+00:00","updated_at":"2026-07-01T01:18:08.128298+00:00"}