{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:D2YMP65HJUGGXBM5L7HDY5MXIR","short_pith_number":"pith:D2YMP65H","schema_version":"1.0","canonical_sha256":"1eb0c7fba74d0c6b859d5fce3c7597444ad32a1ede19776937391c9569e36b72","source":{"kind":"arxiv","id":"2606.01682","version":1},"attestation_state":"computed","paper":{"title":"Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Atoosa Chegini, Soheil Feizi","submitted_at":"2026-06-01T04:43:36Z","abstract_excerpt":"Selecting the best response from multiple small-model samples using a stronger scorer is a simple inference-time strategy, but fails when the small model has already committed to incorrect reasoning paths. PRM guided search avoids this by scoring candidate continuations during generation, but requires a reward model trained with step-level labels.\n  We propose Chunk-Level Guided Generation, a training-free alternative that uses an off-the-shelf large language model as a process scorer. At each step, a small model samples k fixed-length candidate chunks, while the larger model scores the candid"},"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":"2606.01682","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-01T04:43:36Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"8863d21640e629b3675ee6886b3af7fad676de5643267fd790af13ed7eb0f7df","abstract_canon_sha256":"6d6986446094a106897306aef6baf29ffd3a06ad1359f8a7db4f752bfe336ae9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:39.997872Z","signature_b64":"KN200i9f2T4H0scOC9nUPi7lvn+vGdn2xNygu/GMeLFD+f8LB1LKYPLofSGwOWxeW1kNkIW1yBshwZ66ULE+Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1eb0c7fba74d0c6b859d5fce3c7597444ad32a1ede19776937391c9569e36b72","last_reissued_at":"2026-06-02T02:04:39.997533Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:39.997533Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Atoosa Chegini, Soheil Feizi","submitted_at":"2026-06-01T04:43:36Z","abstract_excerpt":"Selecting the best response from multiple small-model samples using a stronger scorer is a simple inference-time strategy, but fails when the small model has already committed to incorrect reasoning paths. PRM guided search avoids this by scoring candidate continuations during generation, but requires a reward model trained with step-level labels.\n  We propose Chunk-Level Guided Generation, a training-free alternative that uses an off-the-shelf large language model as a process scorer. At each step, a small model samples k fixed-length candidate chunks, while the larger model scores the candid"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01682","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/2606.01682/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":"2606.01682","created_at":"2026-06-02T02:04:39.997589+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01682v1","created_at":"2026-06-02T02:04:39.997589+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01682","created_at":"2026-06-02T02:04:39.997589+00:00"},{"alias_kind":"pith_short_12","alias_value":"D2YMP65HJUGG","created_at":"2026-06-02T02:04:39.997589+00:00"},{"alias_kind":"pith_short_16","alias_value":"D2YMP65HJUGGXBM5","created_at":"2026-06-02T02:04:39.997589+00:00"},{"alias_kind":"pith_short_8","alias_value":"D2YMP65H","created_at":"2026-06-02T02:04:39.997589+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/D2YMP65HJUGGXBM5L7HDY5MXIR","json":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR.json","graph_json":"https://pith.science/api/pith-number/D2YMP65HJUGGXBM5L7HDY5MXIR/graph.json","events_json":"https://pith.science/api/pith-number/D2YMP65HJUGGXBM5L7HDY5MXIR/events.json","paper":"https://pith.science/paper/D2YMP65H"},"agent_actions":{"view_html":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR","download_json":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR.json","view_paper":"https://pith.science/paper/D2YMP65H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01682&json=true","fetch_graph":"https://pith.science/api/pith-number/D2YMP65HJUGGXBM5L7HDY5MXIR/graph.json","fetch_events":"https://pith.science/api/pith-number/D2YMP65HJUGGXBM5L7HDY5MXIR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR/action/storage_attestation","attest_author":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR/action/author_attestation","sign_citation":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR/action/citation_signature","submit_replication":"https://pith.science/pith/D2YMP65HJUGGXBM5L7HDY5MXIR/action/replication_record"}},"created_at":"2026-06-02T02:04:39.997589+00:00","updated_at":"2026-06-02T02:04:39.997589+00:00"}