{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:EJO76EXF5I5JF2OJJ5DQI6LG7F","short_pith_number":"pith:EJO76EXF","schema_version":"1.0","canonical_sha256":"225dff12e5ea3a92e9c94f47047966f970563b0a6cbee3883c06ec28c2d24534","source":{"kind":"arxiv","id":"2605.14002","version":1},"attestation_state":"computed","paper":{"title":"PolitNuggets: Benchmarking Agentic Discovery of Long-Tail Political Facts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Current AI agents struggle with fine-grained long-tail political facts and show wide variation in discovery efficiency.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Yifei Zhu","submitted_at":"2026-05-13T18:09:03Z","abstract_excerpt":"Large Reasoning Models (LRMs) embedded in agentic frameworks have transformed information retrieval from static, long context question answering into open-ended exploration. Yet real world use requires models to discover and synthesize \"long-tail\" facts from dispersed sources, a capability that remains under-evaluated. We introduce PolitNuggets, a multilingual benchmark for agentic information synthesis via constructing political biographies for 400 global elites, covering over 10000 political facts. We standardize evaluation with an optimized multi agent system and propose FactNet, an evidenc"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.14002","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T18:09:03Z","cross_cats_sorted":[],"title_canon_sha256":"9765c817aa3899abf0d69aa0966fcb8761149bdea13f8535dcb10bacd89a6b26","abstract_canon_sha256":"00ddd38154ac3174e2f82d6c785709ee23a7b818d352f15b0ed9f59f9d60c7c7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:13.148412Z","signature_b64":"n/S5Zw4KRqLYxhYo/C2pnK8jqVHSRylWIEtXJEldXyde1eJnX+3zEGvYTWqp5BUYEE1efReQ0DBPocxEvF4kAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"225dff12e5ea3a92e9c94f47047966f970563b0a6cbee3883c06ec28c2d24534","last_reissued_at":"2026-05-17T23:39:13.147969Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:13.147969Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PolitNuggets: Benchmarking Agentic Discovery of Long-Tail Political Facts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Current AI agents struggle with fine-grained long-tail political facts and show wide variation in discovery efficiency.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Yifei Zhu","submitted_at":"2026-05-13T18:09:03Z","abstract_excerpt":"Large Reasoning Models (LRMs) embedded in agentic frameworks have transformed information retrieval from static, long context question answering into open-ended exploration. Yet real world use requires models to discover and synthesize \"long-tail\" facts from dispersed sources, a capability that remains under-evaluated. We introduce PolitNuggets, a multilingual benchmark for agentic information synthesis via constructing political biographies for 400 global elites, covering over 10000 political facts. We standardize evaluation with an optimized multi agent system and propose FactNet, an evidenc"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across models and settings, we find that current systems often struggle with fine-grained details, and vary substantially in efficiency. Finally, using benchmark diagnostics, we relate agent performance to underlying model capabilities, highlighting the importance of short-context extraction, multilingual robustness, and reliable tool use.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the over 10,000 political facts assembled for the 400 biographies are accurate ground-truth long-tail information correctly sourced from dispersed sources and that the FactNet protocol validly measures real-world agentic discovery capability.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"PolitNuggets is a multilingual benchmark showing that AI agents struggle with fine-grained accuracy and efficiency when discovering long-tail political facts for elite biographies, linking performance to short-context extraction, multilingual robustness, and tool use.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Current AI agents struggle with fine-grained long-tail political facts and show wide variation in discovery efficiency.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a6889bfdea7d0dd019401348717a8ba3faf2502cca8a67cb61229e799e48a453"},"source":{"id":"2605.14002","kind":"arxiv","version":1},"verdict":{"id":"422fb963-5a7a-418b-97c4-cf6645746d1c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:50:59.086931Z","strongest_claim":"Across models and settings, we find that current systems often struggle with fine-grained details, and vary substantially in efficiency. Finally, using benchmark diagnostics, we relate agent performance to underlying model capabilities, highlighting the importance of short-context extraction, multilingual robustness, and reliable tool use.","one_line_summary":"PolitNuggets is a multilingual benchmark showing that AI agents struggle with fine-grained accuracy and efficiency when discovering long-tail political facts for elite biographies, linking performance to short-context extraction, multilingual robustness, and tool use.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the over 10,000 political facts assembled for the 400 biographies are accurate ground-truth long-tail information correctly sourced from dispersed sources and that the FactNet protocol validly measures real-world agentic discovery capability.","pith_extraction_headline":"Current AI agents struggle with fine-grained long-tail political facts and show wide variation in discovery efficiency."},"references":{"count":22,"sample":[{"doi":"","year":2026,"title":"WebGPT: Browser-assisted question-answering with human feedback","work_id":"e25ef3e1-4848-4cb9-bf28-67a420591165","ref_index":1,"cited_arxiv_id":"2112.09332","is_internal_anchor":true},{"doi":"","year":null,"title":"InThe Twelfth In- ternational Conference on Learning Representations (ICLR)","work_id":"307b261b-ab91-4fe0-abdf-202ebc10b0da","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Consolidated Ground Truth (CGT).The final pooled, evidence-verified biography nuggets for all 400 entities (including the Wikipedia-coverage filter We), which define the evaluation target G and the dy","work_id":"aee83977-87d2-43e7-bb84-6db15b3e71e9","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Cached webpages.The raw retrieved web pages collected during our agentic runs, fix- ing the search snapshot used for all reported numbers and enabling offline re-evaluation","work_id":"ca4141e7-e129-456b-aeb6-be7399651275","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Reason- inginContext","work_id":"a99be689-c97b-48fb-95e2-210aa0cad6d2","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":22,"snapshot_sha256":"37cb473d86ae5039ac3ee9d2478497332b11f2032a0184ef10ba8c0e0775d7ff","internal_anchors":1},"formal_canon":{"evidence_count":3,"snapshot_sha256":"0adee156f346ed1eb8a7d6be858a056438292c870b0565454287cc3b4bdb7f55"},"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":"2605.14002","created_at":"2026-05-17T23:39:13.148036+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.14002v1","created_at":"2026-05-17T23:39:13.148036+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14002","created_at":"2026-05-17T23:39:13.148036+00:00"},{"alias_kind":"pith_short_12","alias_value":"EJO76EXF5I5J","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"EJO76EXF5I5JF2OJ","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"EJO76EXF","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":3,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F","json":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F.json","graph_json":"https://pith.science/api/pith-number/EJO76EXF5I5JF2OJJ5DQI6LG7F/graph.json","events_json":"https://pith.science/api/pith-number/EJO76EXF5I5JF2OJJ5DQI6LG7F/events.json","paper":"https://pith.science/paper/EJO76EXF"},"agent_actions":{"view_html":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F","download_json":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F.json","view_paper":"https://pith.science/paper/EJO76EXF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.14002&json=true","fetch_graph":"https://pith.science/api/pith-number/EJO76EXF5I5JF2OJJ5DQI6LG7F/graph.json","fetch_events":"https://pith.science/api/pith-number/EJO76EXF5I5JF2OJJ5DQI6LG7F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F/action/storage_attestation","attest_author":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F/action/author_attestation","sign_citation":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F/action/citation_signature","submit_replication":"https://pith.science/pith/EJO76EXF5I5JF2OJJ5DQI6LG7F/action/replication_record"}},"created_at":"2026-05-17T23:39:13.148036+00:00","updated_at":"2026-05-17T23:39:13.148036+00:00"}