{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NWOU6MHM4KCGCBFEW7PW5762CI","short_pith_number":"pith:NWOU6MHM","schema_version":"1.0","canonical_sha256":"6d9d4f30ece2846104a4b7df6effda12209e95fb3c14a7f868d4e338e7a01edc","source":{"kind":"arxiv","id":"2605.20740","version":1},"attestation_state":"computed","paper":{"title":"Distribution-Aware Reward: Reinforcement Learning over Predictive Distributions for LLM Regression","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Alan Ritter, Ethan Mendes, Hyungjoo Chae, Jay DeYoung, Jungsoo Park, Varsha Kishore, Wei Xu","submitted_at":"2026-05-20T05:43:40Z","abstract_excerpt":"Large language models can predict real-valued quantities from heterogeneous inputs such as text, code, and molecular strings, but most training objectives score each decoded floating-point number independently, improving point estimates without ensuring calibrated predictive distributions. This limits applications requiring candidate ranking or uncertainty estimation. We introduce Distribution-Aware Reward, an on-policy reinforcement learning objective whose main contribution is to train language models to produce better predictive distributions for regression tasks, rather than only optimizin"},"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":"2605.20740","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T05:43:40Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"ff1bb9af79ccceca54bef0eedfb7c3fe03f639d94b4ca3705c921b0b6584fff7","abstract_canon_sha256":"ca0aa2e60ef6d78088f20b42d4c67eb0b5839af53ed2cd83cc898654195cda16"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:51.672487Z","signature_b64":"yQulVuV+ZLc6JspOj0yTlwbelwrRiVQbRVEgkAKuXfhLKBosxs9/4gD1eAq3xqKTKeXPgd7lUxUxDjfcdjNMDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6d9d4f30ece2846104a4b7df6effda12209e95fb3c14a7f868d4e338e7a01edc","last_reissued_at":"2026-05-21T01:04:51.671849Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:51.671849Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Distribution-Aware Reward: Reinforcement Learning over Predictive Distributions for LLM Regression","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Alan Ritter, Ethan Mendes, Hyungjoo Chae, Jay DeYoung, Jungsoo Park, Varsha Kishore, Wei Xu","submitted_at":"2026-05-20T05:43:40Z","abstract_excerpt":"Large language models can predict real-valued quantities from heterogeneous inputs such as text, code, and molecular strings, but most training objectives score each decoded floating-point number independently, improving point estimates without ensuring calibrated predictive distributions. This limits applications requiring candidate ranking or uncertainty estimation. We introduce Distribution-Aware Reward, an on-policy reinforcement learning objective whose main contribution is to train language models to produce better predictive distributions for regression tasks, rather than only optimizin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20740","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/2605.20740/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":"2605.20740","created_at":"2026-05-21T01:04:51.671979+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.20740v1","created_at":"2026-05-21T01:04:51.671979+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20740","created_at":"2026-05-21T01:04:51.671979+00:00"},{"alias_kind":"pith_short_12","alias_value":"NWOU6MHM4KCG","created_at":"2026-05-21T01:04:51.671979+00:00"},{"alias_kind":"pith_short_16","alias_value":"NWOU6MHM4KCGCBFE","created_at":"2026-05-21T01:04:51.671979+00:00"},{"alias_kind":"pith_short_8","alias_value":"NWOU6MHM","created_at":"2026-05-21T01:04:51.671979+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/NWOU6MHM4KCGCBFEW7PW5762CI","json":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI.json","graph_json":"https://pith.science/api/pith-number/NWOU6MHM4KCGCBFEW7PW5762CI/graph.json","events_json":"https://pith.science/api/pith-number/NWOU6MHM4KCGCBFEW7PW5762CI/events.json","paper":"https://pith.science/paper/NWOU6MHM"},"agent_actions":{"view_html":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI","download_json":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI.json","view_paper":"https://pith.science/paper/NWOU6MHM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.20740&json=true","fetch_graph":"https://pith.science/api/pith-number/NWOU6MHM4KCGCBFEW7PW5762CI/graph.json","fetch_events":"https://pith.science/api/pith-number/NWOU6MHM4KCGCBFEW7PW5762CI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI/action/storage_attestation","attest_author":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI/action/author_attestation","sign_citation":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI/action/citation_signature","submit_replication":"https://pith.science/pith/NWOU6MHM4KCGCBFEW7PW5762CI/action/replication_record"}},"created_at":"2026-05-21T01:04:51.671979+00:00","updated_at":"2026-05-21T01:04:51.671979+00:00"}