{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UDQQVVYTAJJEX5X2OSG2ZGIFHA","short_pith_number":"pith:UDQQVVYT","schema_version":"1.0","canonical_sha256":"a0e10ad71302524bf6fa748dac99053803cc46942443e0890841b9e1763a272a","source":{"kind":"arxiv","id":"2607.09789","version":1},"attestation_state":"computed","paper":{"title":"PHITSBench: an execution-scored benchmark for AI-assisted PHITS radiation-transport input generation using natural language","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cond-mat.mtrl-sci"],"primary_cat":"cs.AI","authors_text":"Svetlana V. Boriskina, Xianglin Ji","submitted_at":"2026-07-08T16:35:22Z","abstract_excerpt":"We introduce PHITSBench, an execution-scored benchmark for the Monte Carlo Particle and Heavy Ion Transport code System (PHITS). PHITSBench comprises 282 transport-scorable tasks spanning three common workflow categories: parameter editing (Edit), syntax repair (Repair ), and complete simulation generation from natural-language descriptions (Reproduce). Each task is evaluated using a Composite Metric Score that combines execution success with agreement between generated and reference transport observables. Using PHITSBench, we evaluate five GPT-5.4-based configurations ranging from zero-shot p"},"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":"2607.09789","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-07-08T16:35:22Z","cross_cats_sorted":["cond-mat.mtrl-sci"],"title_canon_sha256":"861e5851701660ad092cb90c5c7552e50a0b4996962d8bd438e74c0e86a882d1","abstract_canon_sha256":"5b9766a4fbefc395de0aeb3d73515e9e98b386e9e132611b4bd380105f4aa2b0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-14T00:18:35.916862Z","signature_b64":"jUI9zYDR73ChlVFPu3go5a+FF1av75Q5sFFwYSS4CV4qaKVccqNUsk3Mlez3U3LlxZiLqGr0xD0nWr96kyMlAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a0e10ad71302524bf6fa748dac99053803cc46942443e0890841b9e1763a272a","last_reissued_at":"2026-07-14T00:18:35.915970Z","signature_status":"signed_v1","first_computed_at":"2026-07-14T00:18:35.915970Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PHITSBench: an execution-scored benchmark for AI-assisted PHITS radiation-transport input generation using natural language","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cond-mat.mtrl-sci"],"primary_cat":"cs.AI","authors_text":"Svetlana V. Boriskina, Xianglin Ji","submitted_at":"2026-07-08T16:35:22Z","abstract_excerpt":"We introduce PHITSBench, an execution-scored benchmark for the Monte Carlo Particle and Heavy Ion Transport code System (PHITS). PHITSBench comprises 282 transport-scorable tasks spanning three common workflow categories: parameter editing (Edit), syntax repair (Repair ), and complete simulation generation from natural-language descriptions (Reproduce). Each task is evaluated using a Composite Metric Score that combines execution success with agreement between generated and reference transport observables. Using PHITSBench, we evaluate five GPT-5.4-based configurations ranging from zero-shot p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.09789","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/2607.09789/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":"2607.09789","created_at":"2026-07-14T00:18:35.916419+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.09789v1","created_at":"2026-07-14T00:18:35.916419+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.09789","created_at":"2026-07-14T00:18:35.916419+00:00"},{"alias_kind":"pith_short_12","alias_value":"UDQQVVYTAJJE","created_at":"2026-07-14T00:18:35.916419+00:00"},{"alias_kind":"pith_short_16","alias_value":"UDQQVVYTAJJEX5X2","created_at":"2026-07-14T00:18:35.916419+00:00"},{"alias_kind":"pith_short_8","alias_value":"UDQQVVYT","created_at":"2026-07-14T00:18:35.916419+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/UDQQVVYTAJJEX5X2OSG2ZGIFHA","json":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA.json","graph_json":"https://pith.science/api/pith-number/UDQQVVYTAJJEX5X2OSG2ZGIFHA/graph.json","events_json":"https://pith.science/api/pith-number/UDQQVVYTAJJEX5X2OSG2ZGIFHA/events.json","paper":"https://pith.science/paper/UDQQVVYT"},"agent_actions":{"view_html":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA","download_json":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA.json","view_paper":"https://pith.science/paper/UDQQVVYT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.09789&json=true","fetch_graph":"https://pith.science/api/pith-number/UDQQVVYTAJJEX5X2OSG2ZGIFHA/graph.json","fetch_events":"https://pith.science/api/pith-number/UDQQVVYTAJJEX5X2OSG2ZGIFHA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA/action/storage_attestation","attest_author":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA/action/author_attestation","sign_citation":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA/action/citation_signature","submit_replication":"https://pith.science/pith/UDQQVVYTAJJEX5X2OSG2ZGIFHA/action/replication_record"}},"created_at":"2026-07-14T00:18:35.916419+00:00","updated_at":"2026-07-14T00:18:35.916419+00:00"}