{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XUPCXYLYCI2ADMWSND3PLOPIQI","short_pith_number":"pith:XUPCXYLY","schema_version":"1.0","canonical_sha256":"bd1e2be178123401b2d268f6f5b9e8820d5666933f5efca532ec26c1de6c224a","source":{"kind":"arxiv","id":"2507.18618","version":1},"attestation_state":"computed","paper":{"title":"TRPrompt: Bootstrapping Query-Aware Prompt Optimization from Textual Rewards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Andreea Nica, Ivan Zakazov, Nicolas Mario Baldwin, Robert West, Saibo Geng","submitted_at":"2025-07-24T17:54:44Z","abstract_excerpt":"Prompt optimization improves the reasoning abilities of large language models (LLMs) without requiring parameter updates to the target model. Following heuristic-based \"Think step by step\" approaches, the field has evolved in two main directions: while one group of methods uses textual feedback to elicit improved prompts from general-purpose LLMs in a training-free way, a concurrent line of research relies on numerical rewards to train a special prompt model, tailored for providing optimal prompts to the target model. In this paper, we introduce the Textual Reward Prompt framework (TRPrompt), "},"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":"2507.18618","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-07-24T17:54:44Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2890f729f0139fa184e9e3db3d01a0d16513ec920b35e9c39ea009221e746ecf","abstract_canon_sha256":"1a3a621b811ac430b7ee2c6b3b09d086f19af99fc2ea4ad8af47e7af162d876b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:42:51.159269Z","signature_b64":"mEMtJCR7ObkaG4Yas4+R6wW2FXErydv37RQKyRT5oj4Zd4fjbqU5NZ+m/RTYqGAJFrSmirAK8AmA8Id+DA9iCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bd1e2be178123401b2d268f6f5b9e8820d5666933f5efca532ec26c1de6c224a","last_reissued_at":"2026-07-05T11:42:51.158774Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:42:51.158774Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TRPrompt: Bootstrapping Query-Aware Prompt Optimization from Textual Rewards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Andreea Nica, Ivan Zakazov, Nicolas Mario Baldwin, Robert West, Saibo Geng","submitted_at":"2025-07-24T17:54:44Z","abstract_excerpt":"Prompt optimization improves the reasoning abilities of large language models (LLMs) without requiring parameter updates to the target model. Following heuristic-based \"Think step by step\" approaches, the field has evolved in two main directions: while one group of methods uses textual feedback to elicit improved prompts from general-purpose LLMs in a training-free way, a concurrent line of research relies on numerical rewards to train a special prompt model, tailored for providing optimal prompts to the target model. In this paper, we introduce the Textual Reward Prompt framework (TRPrompt), "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.18618","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/2507.18618/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":"2507.18618","created_at":"2026-07-05T11:42:51.158833+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.18618v1","created_at":"2026-07-05T11:42:51.158833+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.18618","created_at":"2026-07-05T11:42:51.158833+00:00"},{"alias_kind":"pith_short_12","alias_value":"XUPCXYLYCI2A","created_at":"2026-07-05T11:42:51.158833+00:00"},{"alias_kind":"pith_short_16","alias_value":"XUPCXYLYCI2ADMWS","created_at":"2026-07-05T11:42:51.158833+00:00"},{"alias_kind":"pith_short_8","alias_value":"XUPCXYLY","created_at":"2026-07-05T11:42:51.158833+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2511.01016","citing_title":"Prompt-R1: Collaborative Automatic Prompting Framework via End-to-end Reinforcement Learning","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI","json":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI.json","graph_json":"https://pith.science/api/pith-number/XUPCXYLYCI2ADMWSND3PLOPIQI/graph.json","events_json":"https://pith.science/api/pith-number/XUPCXYLYCI2ADMWSND3PLOPIQI/events.json","paper":"https://pith.science/paper/XUPCXYLY"},"agent_actions":{"view_html":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI","download_json":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI.json","view_paper":"https://pith.science/paper/XUPCXYLY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.18618&json=true","fetch_graph":"https://pith.science/api/pith-number/XUPCXYLYCI2ADMWSND3PLOPIQI/graph.json","fetch_events":"https://pith.science/api/pith-number/XUPCXYLYCI2ADMWSND3PLOPIQI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI/action/storage_attestation","attest_author":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI/action/author_attestation","sign_citation":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI/action/citation_signature","submit_replication":"https://pith.science/pith/XUPCXYLYCI2ADMWSND3PLOPIQI/action/replication_record"}},"created_at":"2026-07-05T11:42:51.158833+00:00","updated_at":"2026-07-05T11:42:51.158833+00:00"}