{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3RMZBBWNSCAI2FKV7FYEQM3SNB","short_pith_number":"pith:3RMZBBWN","schema_version":"1.0","canonical_sha256":"dc599086cd90808d1555f970483372686b35747649a8ea878fb6193782afa961","source":{"kind":"arxiv","id":"2408.13001","version":2},"attestation_state":"computed","paper":{"title":"CRUXEval-X: A Benchmark for Multilingual Code Reasoning, Understanding and Execution","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Ben He, Hongyu Lin, Jialun Cao, Le Sun, Ming Wen, Ruiyang Xu, Shing-Chi Cheung, Xianpei Han, Yaojie Lu","submitted_at":"2024-08-23T11:43:00Z","abstract_excerpt":"Code benchmarks such as HumanEval are widely adopted to evaluate Large Language Models' (LLMs) coding capabilities. However, there is an unignorable programming language bias in existing code benchmarks -- over 95% code generation benchmarks are dominated by Python, leaving the LLMs' capabilities in other programming languages such as Java and C/C++ unknown. Moreover, coding task bias is also crucial. Most benchmarks focus on code generation capability, while benchmarks for code reasoning (given input, reasoning output; and given output, reasoning input), an essential coding capability, are in"},"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":"2408.13001","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-08-23T11:43:00Z","cross_cats_sorted":[],"title_canon_sha256":"28cb97089ddad76ce55ec3fe27677151a980c172756b8f86c7b83c48f628c68e","abstract_canon_sha256":"280978558d4cf3cebcc1a2d3a7ba94f2fe15f632cd6dae9ab7f71dabbdb21e05"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:04:18.360817Z","signature_b64":"loyR7WDHHA5YPvfwKsIusEHekjP1qq1IF2nkBB087pRO38bidONGGziFENJAIOcigWBWC3bpNHcGbDSwXt+hCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dc599086cd90808d1555f970483372686b35747649a8ea878fb6193782afa961","last_reissued_at":"2026-07-05T11:04:18.360265Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:04:18.360265Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CRUXEval-X: A Benchmark for Multilingual Code Reasoning, Understanding and Execution","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Ben He, Hongyu Lin, Jialun Cao, Le Sun, Ming Wen, Ruiyang Xu, Shing-Chi Cheung, Xianpei Han, Yaojie Lu","submitted_at":"2024-08-23T11:43:00Z","abstract_excerpt":"Code benchmarks such as HumanEval are widely adopted to evaluate Large Language Models' (LLMs) coding capabilities. However, there is an unignorable programming language bias in existing code benchmarks -- over 95% code generation benchmarks are dominated by Python, leaving the LLMs' capabilities in other programming languages such as Java and C/C++ unknown. Moreover, coding task bias is also crucial. Most benchmarks focus on code generation capability, while benchmarks for code reasoning (given input, reasoning output; and given output, reasoning input), an essential coding capability, are in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.13001","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/2408.13001/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":"2408.13001","created_at":"2026-07-05T11:04:18.360337+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.13001v2","created_at":"2026-07-05T11:04:18.360337+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.13001","created_at":"2026-07-05T11:04:18.360337+00:00"},{"alias_kind":"pith_short_12","alias_value":"3RMZBBWNSCAI","created_at":"2026-07-05T11:04:18.360337+00:00"},{"alias_kind":"pith_short_16","alias_value":"3RMZBBWNSCAI2FKV","created_at":"2026-07-05T11:04:18.360337+00:00"},{"alias_kind":"pith_short_8","alias_value":"3RMZBBWN","created_at":"2026-07-05T11:04:18.360337+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.31725","citing_title":"Do Machines Struggle Where Humans Do? LLM and Human Comprehension of Obfuscated Code","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30182","citing_title":"MirrorCode: AI can rebuild entire programs from behavior alone","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17958","citing_title":"Enhancing the Code Reasoning Capabilities of LLMs via Consistency-based Reinforcement Learning","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2506.03535","citing_title":"Across Programming Language Silos: A Study on Cross-Lingual Retrieval-augmented Code Generation","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11922","citing_title":"StepCodeReasoner: Aligning Code Reasoning with Stepwise Execution Traces via Reinforcement Learning","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB","json":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB.json","graph_json":"https://pith.science/api/pith-number/3RMZBBWNSCAI2FKV7FYEQM3SNB/graph.json","events_json":"https://pith.science/api/pith-number/3RMZBBWNSCAI2FKV7FYEQM3SNB/events.json","paper":"https://pith.science/paper/3RMZBBWN"},"agent_actions":{"view_html":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB","download_json":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB.json","view_paper":"https://pith.science/paper/3RMZBBWN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.13001&json=true","fetch_graph":"https://pith.science/api/pith-number/3RMZBBWNSCAI2FKV7FYEQM3SNB/graph.json","fetch_events":"https://pith.science/api/pith-number/3RMZBBWNSCAI2FKV7FYEQM3SNB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB/action/storage_attestation","attest_author":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB/action/author_attestation","sign_citation":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB/action/citation_signature","submit_replication":"https://pith.science/pith/3RMZBBWNSCAI2FKV7FYEQM3SNB/action/replication_record"}},"created_at":"2026-07-05T11:04:18.360337+00:00","updated_at":"2026-07-05T11:04:18.360337+00:00"}