{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:OUEP6ODUYVUO64TJOAWPPD3ODL","short_pith_number":"pith:OUEP6ODU","canonical_record":{"source":{"id":"2605.19102","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-18T20:42:23Z","cross_cats_sorted":[],"title_canon_sha256":"6bd51326c1b33480a69770533d3664c8b40b307c6e7e39d6949114c715b6bd1f","abstract_canon_sha256":"2cec96d18416afb994c20839512fc927d6845a67ab5c3893a6654a8232ada203"},"schema_version":"1.0"},"canonical_sha256":"7508ff3874c568ef7269702cf78f6e1ac1bfa16a1c1812f1efda0454b9cf8b47","source":{"kind":"arxiv","id":"2605.19102","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.19102","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"arxiv_version","alias_value":"2605.19102v1","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19102","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"pith_short_12","alias_value":"OUEP6ODUYVUO","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"pith_short_16","alias_value":"OUEP6ODUYVUO64TJ","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"pith_short_8","alias_value":"OUEP6ODU","created_at":"2026-05-20T01:05:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:OUEP6ODUYVUO64TJOAWPPD3ODL","target":"record","payload":{"canonical_record":{"source":{"id":"2605.19102","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-18T20:42:23Z","cross_cats_sorted":[],"title_canon_sha256":"6bd51326c1b33480a69770533d3664c8b40b307c6e7e39d6949114c715b6bd1f","abstract_canon_sha256":"2cec96d18416afb994c20839512fc927d6845a67ab5c3893a6654a8232ada203"},"schema_version":"1.0"},"canonical_sha256":"7508ff3874c568ef7269702cf78f6e1ac1bfa16a1c1812f1efda0454b9cf8b47","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:27.230066Z","signature_b64":"uARfLUS2402itBrnuAsjYnO/hQ2erywW9cJxSrk5oJxXKASgbRRhPsMNjnlUspe8ht7L1P7yZ1g7AIhByPjLDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7508ff3874c568ef7269702cf78f6e1ac1bfa16a1c1812f1efda0454b9cf8b47","last_reissued_at":"2026-05-20T01:05:27.229250Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:27.229250Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.19102","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:05:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cNlkv90shZUQCCPLVlASeRhfmmsgV+WS31dSdOL0G9CGpdu6NXq69BvvcnpYLhV1Q7ALlvnHTWXyaBod2Y7SDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T12:01:33.821984Z"},"content_sha256":"8c17fa844959ccef2fb9622b3896a8a599fca11c8048df736223700ac9a803a7","schema_version":"1.0","event_id":"sha256:8c17fa844959ccef2fb9622b3896a8a599fca11c8048df736223700ac9a803a7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:OUEP6ODUYVUO64TJOAWPPD3ODL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Prompt Optimization for LLM Code Generation via Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Ali Mohammadi Esfahani, Nafiseh Kahani, Samuel A.Ajila","submitted_at":"2026-05-18T20:42:23Z","abstract_excerpt":"Large Language Models (LLMs) can generate code from natural language, but their performance is highly sensitive to prompt formulation. We propose a reinforcement-learning-based framework that models prompt refinement as a sequential decision-making problem. A Proximal Policy Optimization (PPO) agent iteratively improves prompts using a hybrid action space that combines direct generation, genetic lexical mutation and semantic rewriting, guided by shaped rewards derived from unit-test feedback. We evaluate the framework on MBPP+, HumanEval+, and APPS using CodeT5+, CodeLLaMA, and DeepSeek-Coder "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19102","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/2605.19102/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:05:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7bKtaN5UdNroF0sfZ4t07NCS+WUyyZaxVBtRcPWxWnKbch899+G/3z5l4CFCPh+0NMwRAy0/HBu/0ie1SbKKBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T12:01:33.822705Z"},"content_sha256":"6d2394bc7e5413cd98456c7df811312690fe953e7b4a121d5614f1ab1f1a80e9","schema_version":"1.0","event_id":"sha256:6d2394bc7e5413cd98456c7df811312690fe953e7b4a121d5614f1ab1f1a80e9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OUEP6ODUYVUO64TJOAWPPD3ODL/bundle.json","state_url":"https://pith.science/pith/OUEP6ODUYVUO64TJOAWPPD3ODL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OUEP6ODUYVUO64TJOAWPPD3ODL/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-25T12:01:33Z","links":{"resolver":"https://pith.science/pith/OUEP6ODUYVUO64TJOAWPPD3ODL","bundle":"https://pith.science/pith/OUEP6ODUYVUO64TJOAWPPD3ODL/bundle.json","state":"https://pith.science/pith/OUEP6ODUYVUO64TJOAWPPD3ODL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OUEP6ODUYVUO64TJOAWPPD3ODL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:OUEP6ODUYVUO64TJOAWPPD3ODL","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"2cec96d18416afb994c20839512fc927d6845a67ab5c3893a6654a8232ada203","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-18T20:42:23Z","title_canon_sha256":"6bd51326c1b33480a69770533d3664c8b40b307c6e7e39d6949114c715b6bd1f"},"schema_version":"1.0","source":{"id":"2605.19102","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.19102","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"arxiv_version","alias_value":"2605.19102v1","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19102","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"pith_short_12","alias_value":"OUEP6ODUYVUO","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"pith_short_16","alias_value":"OUEP6ODUYVUO64TJ","created_at":"2026-05-20T01:05:27Z"},{"alias_kind":"pith_short_8","alias_value":"OUEP6ODU","created_at":"2026-05-20T01:05:27Z"}],"graph_snapshots":[{"event_id":"sha256:6d2394bc7e5413cd98456c7df811312690fe953e7b4a121d5614f1ab1f1a80e9","target":"graph","created_at":"2026-05-20T01:05:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2605.19102/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) can generate code from natural language, but their performance is highly sensitive to prompt formulation. We propose a reinforcement-learning-based framework that models prompt refinement as a sequential decision-making problem. A Proximal Policy Optimization (PPO) agent iteratively improves prompts using a hybrid action space that combines direct generation, genetic lexical mutation and semantic rewriting, guided by shaped rewards derived from unit-test feedback. We evaluate the framework on MBPP+, HumanEval+, and APPS using CodeT5+, CodeLLaMA, and DeepSeek-Coder ","authors_text":"Ali Mohammadi Esfahani, Nafiseh Kahani, Samuel A.Ajila","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-18T20:42:23Z","title":"Prompt Optimization for LLM Code Generation via Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19102","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:8c17fa844959ccef2fb9622b3896a8a599fca11c8048df736223700ac9a803a7","target":"record","created_at":"2026-05-20T01:05:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"2cec96d18416afb994c20839512fc927d6845a67ab5c3893a6654a8232ada203","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-18T20:42:23Z","title_canon_sha256":"6bd51326c1b33480a69770533d3664c8b40b307c6e7e39d6949114c715b6bd1f"},"schema_version":"1.0","source":{"id":"2605.19102","kind":"arxiv","version":1}},"canonical_sha256":"7508ff3874c568ef7269702cf78f6e1ac1bfa16a1c1812f1efda0454b9cf8b47","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7508ff3874c568ef7269702cf78f6e1ac1bfa16a1c1812f1efda0454b9cf8b47","first_computed_at":"2026-05-20T01:05:27.229250Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T01:05:27.229250Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"uARfLUS2402itBrnuAsjYnO/hQ2erywW9cJxSrk5oJxXKASgbRRhPsMNjnlUspe8ht7L1P7yZ1g7AIhByPjLDQ==","signature_status":"signed_v1","signed_at":"2026-05-20T01:05:27.230066Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.19102","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8c17fa844959ccef2fb9622b3896a8a599fca11c8048df736223700ac9a803a7","sha256:6d2394bc7e5413cd98456c7df811312690fe953e7b4a121d5614f1ab1f1a80e9"],"state_sha256":"5af9c6508bd3cf3253fb5847f7d96e4321bbee21ef1aaaf158fd68bf0c7fb5ca"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nhiSyfYTEs4RyZ8GhqV8XEpr53fGy5Jxky9RQ4De3XWGbmBpJC36xuUciZ2XwD07v6BcBlgvxurlMq7EAHBbCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T12:01:33.827229Z","bundle_sha256":"519a013283f3a288b5ade9972fba4347a9b91ebc521504dc1c9fd11b03c6620f"}}