{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:7VR3TNGMFNCPHL5JH722NIYTUP","short_pith_number":"pith:7VR3TNGM","schema_version":"1.0","canonical_sha256":"fd63b9b4cc2b44f3afa93ff5a6a313a3fed39ff5ca1783b76f09f64dc5672b89","source":{"kind":"arxiv","id":"2405.10276","version":2},"attestation_state":"computed","paper":{"title":"Revisiting OPRO: The Limitations of Small-Scale LLMs as Optimizers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.HC"],"primary_cat":"cs.CL","authors_text":"Jinyue Yuan, Salman Avestimehr, Tuo Zhang","submitted_at":"2024-05-16T17:33:50Z","abstract_excerpt":"Numerous recent works aim to enhance the efficacy of Large Language Models (LLMs) through strategic prompting. In particular, the Optimization by PROmpting (OPRO) approach provides state-of-the-art performance by leveraging LLMs as optimizers where the optimization task is to find instructions that maximize the task accuracy. In this paper, we revisit OPRO for automated prompting with relatively small-scale LLMs, such as LLaMa-2 family and Mistral 7B. Our investigation reveals that OPRO shows limited effectiveness in small-scale LLMs, with limited inference capabilities constraining optimizati"},"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":"2405.10276","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-16T17:33:50Z","cross_cats_sorted":["cs.HC"],"title_canon_sha256":"06e434f0f2d8d73412ff625d00635509a55cfa18fa03434d3a5294f8979516a6","abstract_canon_sha256":"31897a9be06d7891f095f8a392e3c404f2fb6eb5f814b440021fc7b004fa2bdd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:45:51.543669Z","signature_b64":"8Myz1CjEBvgTTCnmHNfNbs/xSfj/cva7dT0H3SOE8xSTLEXu4GMZBQxj3yz0jh/nFRFln5lmhsAz+mk4yuAKCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fd63b9b4cc2b44f3afa93ff5a6a313a3fed39ff5ca1783b76f09f64dc5672b89","last_reissued_at":"2026-07-05T08:45:51.543246Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:45:51.543246Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Revisiting OPRO: The Limitations of Small-Scale LLMs as Optimizers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.HC"],"primary_cat":"cs.CL","authors_text":"Jinyue Yuan, Salman Avestimehr, Tuo Zhang","submitted_at":"2024-05-16T17:33:50Z","abstract_excerpt":"Numerous recent works aim to enhance the efficacy of Large Language Models (LLMs) through strategic prompting. In particular, the Optimization by PROmpting (OPRO) approach provides state-of-the-art performance by leveraging LLMs as optimizers where the optimization task is to find instructions that maximize the task accuracy. In this paper, we revisit OPRO for automated prompting with relatively small-scale LLMs, such as LLaMa-2 family and Mistral 7B. Our investigation reveals that OPRO shows limited effectiveness in small-scale LLMs, with limited inference capabilities constraining optimizati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.10276","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/2405.10276/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":"2405.10276","created_at":"2026-07-05T08:45:51.543302+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.10276v2","created_at":"2026-07-05T08:45:51.543302+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.10276","created_at":"2026-07-05T08:45:51.543302+00:00"},{"alias_kind":"pith_short_12","alias_value":"7VR3TNGMFNCP","created_at":"2026-07-05T08:45:51.543302+00:00"},{"alias_kind":"pith_short_16","alias_value":"7VR3TNGMFNCPHL5J","created_at":"2026-07-05T08:45:51.543302+00:00"},{"alias_kind":"pith_short_8","alias_value":"7VR3TNGM","created_at":"2026-07-05T08:45:51.543302+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21641","citing_title":"When Is an LLM Worth It for Hyperparameter Optimization? A Budget-Matched Study on Tabular Data Finds the Warm-Start Is a Default Configuration, Not the Model","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21641","citing_title":"When Is an LLM Worth It for Hyperparameter Optimization? A Budget-Matched Study on Tabular Data Finds the Warm-Start Is a Default Configuration, Not the Model","ref_index":44,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP","json":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP.json","graph_json":"https://pith.science/api/pith-number/7VR3TNGMFNCPHL5JH722NIYTUP/graph.json","events_json":"https://pith.science/api/pith-number/7VR3TNGMFNCPHL5JH722NIYTUP/events.json","paper":"https://pith.science/paper/7VR3TNGM"},"agent_actions":{"view_html":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP","download_json":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP.json","view_paper":"https://pith.science/paper/7VR3TNGM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.10276&json=true","fetch_graph":"https://pith.science/api/pith-number/7VR3TNGMFNCPHL5JH722NIYTUP/graph.json","fetch_events":"https://pith.science/api/pith-number/7VR3TNGMFNCPHL5JH722NIYTUP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP/action/storage_attestation","attest_author":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP/action/author_attestation","sign_citation":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP/action/citation_signature","submit_replication":"https://pith.science/pith/7VR3TNGMFNCPHL5JH722NIYTUP/action/replication_record"}},"created_at":"2026-07-05T08:45:51.543302+00:00","updated_at":"2026-07-05T08:45:51.543302+00:00"}