{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SUQX6L7CL4KGODCAJ5RZN2PIDF","short_pith_number":"pith:SUQX6L7C","schema_version":"1.0","canonical_sha256":"95217f2fe25f14670c404f6396e9e81972f035c10ed2e347359b7e73eaa481ad","source":{"kind":"arxiv","id":"2603.18444","version":2},"attestation_state":"computed","paper":{"title":"Discounted Beta-Bernoulli Reward Estimation for Sample-Efficient Reinforcement Learning with Verifiable Rewards","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Doohyuk Jang, Eunho Yang, Gyouk Chu, Haechan Kim, Soohyun Ryu","submitted_at":"2026-03-19T03:10:47Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) has emerged as an effective post-training paradigm for improving the reasoning capabilities of large language models. However, existing group-based RLVR methods often suffer from severe sample inefficiency. This inefficiency stems from reliance on point estimation of rewards from a small number of rollouts, leading to high estimation variance, variance collapse, and ineffective utilization of generated responses. In this work, we reformulate RLVR from a statistical estimation perspective by modeling rewards as samples drawn from a policy-in"},"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":"2603.18444","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-03-19T03:10:47Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7e132d2705964b63c1be48a6ad81d7f6e4120cd5e19ea24e5912c3208b197c25","abstract_canon_sha256":"25e063ea992fc30a8f55c19f4b195f97b0b41372d1072154912222e20d25f838"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:28.376092Z","signature_b64":"y5vsWw260IMZCwVUXgCxDgotqpMN+OhNUl1dOjA9XkasC5JDwzCM+84accc5YPl7vWtnDdIYcQr513WZUyA+BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"95217f2fe25f14670c404f6396e9e81972f035c10ed2e347359b7e73eaa481ad","last_reissued_at":"2026-05-26T01:03:28.375177Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:28.375177Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Discounted Beta-Bernoulli Reward Estimation for Sample-Efficient Reinforcement Learning with Verifiable Rewards","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Doohyuk Jang, Eunho Yang, Gyouk Chu, Haechan Kim, Soohyun Ryu","submitted_at":"2026-03-19T03:10:47Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) has emerged as an effective post-training paradigm for improving the reasoning capabilities of large language models. However, existing group-based RLVR methods often suffer from severe sample inefficiency. This inefficiency stems from reliance on point estimation of rewards from a small number of rollouts, leading to high estimation variance, variance collapse, and ineffective utilization of generated responses. In this work, we reformulate RLVR from a statistical estimation perspective by modeling rewards as samples drawn from a policy-in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.18444","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/2603.18444/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":"2603.18444","created_at":"2026-05-26T01:03:28.375306+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.18444v2","created_at":"2026-05-26T01:03:28.375306+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.18444","created_at":"2026-05-26T01:03:28.375306+00:00"},{"alias_kind":"pith_short_12","alias_value":"SUQX6L7CL4KG","created_at":"2026-05-26T01:03:28.375306+00:00"},{"alias_kind":"pith_short_16","alias_value":"SUQX6L7CL4KGODCA","created_at":"2026-05-26T01:03:28.375306+00:00"},{"alias_kind":"pith_short_8","alias_value":"SUQX6L7C","created_at":"2026-05-26T01:03:28.375306+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/SUQX6L7CL4KGODCAJ5RZN2PIDF","json":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF.json","graph_json":"https://pith.science/api/pith-number/SUQX6L7CL4KGODCAJ5RZN2PIDF/graph.json","events_json":"https://pith.science/api/pith-number/SUQX6L7CL4KGODCAJ5RZN2PIDF/events.json","paper":"https://pith.science/paper/SUQX6L7C"},"agent_actions":{"view_html":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF","download_json":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF.json","view_paper":"https://pith.science/paper/SUQX6L7C","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.18444&json=true","fetch_graph":"https://pith.science/api/pith-number/SUQX6L7CL4KGODCAJ5RZN2PIDF/graph.json","fetch_events":"https://pith.science/api/pith-number/SUQX6L7CL4KGODCAJ5RZN2PIDF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF/action/storage_attestation","attest_author":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF/action/author_attestation","sign_citation":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF/action/citation_signature","submit_replication":"https://pith.science/pith/SUQX6L7CL4KGODCAJ5RZN2PIDF/action/replication_record"}},"created_at":"2026-05-26T01:03:28.375306+00:00","updated_at":"2026-05-26T01:03:28.375306+00:00"}