{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:MYW6MOCPH4ES2VBEPSJY77UPJV","short_pith_number":"pith:MYW6MOCP","schema_version":"1.0","canonical_sha256":"662de6384f3f092d54247c938ffe8f4d765ee8f7b555e13af23758933eef225d","source":{"kind":"arxiv","id":"2406.10248","version":4},"attestation_state":"computed","paper":{"title":"On the Worst Prompt Performance of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bowen Cao, Deng Cai, Wai Lam, Yuexian Zou, Zhisong Zhang","submitted_at":"2024-06-08T13:40:38Z","abstract_excerpt":"The performance of large language models (LLMs) is acutely sensitive to the phrasing of prompts, which raises significant concerns about their reliability in real-world scenarios. Existing studies often divide prompts into task-level instructions and case-level inputs and primarily focus on evaluating and improving robustness against variations in tasks-level instructions. However, this setup fails to fully address the diversity of real-world user queries and assumes the existence of task-specific datasets. To address these limitations, we introduce RobustAlpacaEval, a new benchmark that consi"},"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":"2406.10248","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-08T13:40:38Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d5cd2233c4b9b12c3e70a03a5fbbf035f127c38b226af0f7faf82a7a747e1395","abstract_canon_sha256":"78a467d8c27d407cdf21b260355675660ad028320d81af3eb5094952b675161f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:28:32.042478Z","signature_b64":"zxbcn5jTAN7/5mSSwi+IFNtQoKe1QvhKKJ5tQsNWh55Z4ggxcEfqijMSZriPHwD4rFMVyd6x0wbcImb8n0X1CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"662de6384f3f092d54247c938ffe8f4d765ee8f7b555e13af23758933eef225d","last_reissued_at":"2026-07-05T09:28:32.041931Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:28:32.041931Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Worst Prompt Performance of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bowen Cao, Deng Cai, Wai Lam, Yuexian Zou, Zhisong Zhang","submitted_at":"2024-06-08T13:40:38Z","abstract_excerpt":"The performance of large language models (LLMs) is acutely sensitive to the phrasing of prompts, which raises significant concerns about their reliability in real-world scenarios. Existing studies often divide prompts into task-level instructions and case-level inputs and primarily focus on evaluating and improving robustness against variations in tasks-level instructions. However, this setup fails to fully address the diversity of real-world user queries and assumes the existence of task-specific datasets. To address these limitations, we introduce RobustAlpacaEval, a new benchmark that consi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.10248","kind":"arxiv","version":4},"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/2406.10248/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":"2406.10248","created_at":"2026-07-05T09:28:32.041986+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.10248v4","created_at":"2026-07-05T09:28:32.041986+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.10248","created_at":"2026-07-05T09:28:32.041986+00:00"},{"alias_kind":"pith_short_12","alias_value":"MYW6MOCPH4ES","created_at":"2026-07-05T09:28:32.041986+00:00"},{"alias_kind":"pith_short_16","alias_value":"MYW6MOCPH4ES2VBE","created_at":"2026-07-05T09:28:32.041986+00:00"},{"alias_kind":"pith_short_8","alias_value":"MYW6MOCP","created_at":"2026-07-05T09:28:32.041986+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22797","citing_title":"Measuring Behavior Portability in Large Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2508.15411","citing_title":"Foundational Design Principles and Patterns for Building Robust and Adaptive GenAI-Native Systems","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01048","citing_title":"Compared to What? Baselines and Metrics for Counterfactual Prompting","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV","json":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV.json","graph_json":"https://pith.science/api/pith-number/MYW6MOCPH4ES2VBEPSJY77UPJV/graph.json","events_json":"https://pith.science/api/pith-number/MYW6MOCPH4ES2VBEPSJY77UPJV/events.json","paper":"https://pith.science/paper/MYW6MOCP"},"agent_actions":{"view_html":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV","download_json":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV.json","view_paper":"https://pith.science/paper/MYW6MOCP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.10248&json=true","fetch_graph":"https://pith.science/api/pith-number/MYW6MOCPH4ES2VBEPSJY77UPJV/graph.json","fetch_events":"https://pith.science/api/pith-number/MYW6MOCPH4ES2VBEPSJY77UPJV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV/action/storage_attestation","attest_author":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV/action/author_attestation","sign_citation":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV/action/citation_signature","submit_replication":"https://pith.science/pith/MYW6MOCPH4ES2VBEPSJY77UPJV/action/replication_record"}},"created_at":"2026-07-05T09:28:32.041986+00:00","updated_at":"2026-07-05T09:28:32.041986+00:00"}