{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LMDI7SLNZY7VBQ2NXJW2ANINAD","short_pith_number":"pith:LMDI7SLN","schema_version":"1.0","canonical_sha256":"5b068fc96dce3f50c34dba6da0350d00c08991f849e6e0776f17a7a5755212ea","source":{"kind":"arxiv","id":"2512.00417","version":5},"attestation_state":"computed","paper":{"title":"CryptoBench: A Dynamic Benchmark for Expert-Level Evaluation of LLM Agents in Cryptocurrency","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chaoyang He, Jason Ge, Jiacheng Guo, Jiashuo Liu, Jia Tian, Kaixuan Huang, Kanghong Zhan, Lin Yang, Mengdi Wang, Nicholas Deng, Qixin Xiao, Suozhi Huang, Tianyi Li, Wenhao Huang, Xiaochen Liu, Yifan Zhang, Yifu Lu, Zihao Li, Zixin Yao","submitted_at":"2025-11-29T09:52:34Z","abstract_excerpt":"This paper introduces CryptoBench, the first expert-curated, dynamic benchmark designed to rigorously evaluate the real-world capabilities of Large Language Model (LLM) agents in the uniquely demanding and fast-paced cryptocurrency domain. Unlike general-purpose agent benchmarks for search and prediction, professional crypto analysis presents specific challenges: \\emph{extreme time-sensitivity}, \\emph{a highly adversarial information environment}, and the critical need to synthesize data from \\emph{diverse, specialized sources}, such as on-chain intelligence platforms and real-time Decentraliz"},"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":"2512.00417","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-11-29T09:52:34Z","cross_cats_sorted":[],"title_canon_sha256":"e2e38a973ea0fec9a987ac2b8d9ef9c604ff1f00fd3ccc8874d955d8ba677c96","abstract_canon_sha256":"bed023db3b169e785274285c3446c9c1ac5e9245dda700fdeb885a033e3b60a7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:01:37.435964Z","signature_b64":"is4mKaNA4cDdIN3cAc56uoD6YkRCxvdn/EUlhlvV6OCenItM/sUl59ZUI8HqQR1Fm/ob4tqEq2Ha7vUPoXhOAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b068fc96dce3f50c34dba6da0350d00c08991f849e6e0776f17a7a5755212ea","last_reissued_at":"2026-05-20T00:01:37.435398Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:01:37.435398Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CryptoBench: A Dynamic Benchmark for Expert-Level Evaluation of LLM Agents in Cryptocurrency","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chaoyang He, Jason Ge, Jiacheng Guo, Jiashuo Liu, Jia Tian, Kaixuan Huang, Kanghong Zhan, Lin Yang, Mengdi Wang, Nicholas Deng, Qixin Xiao, Suozhi Huang, Tianyi Li, Wenhao Huang, Xiaochen Liu, Yifan Zhang, Yifu Lu, Zihao Li, Zixin Yao","submitted_at":"2025-11-29T09:52:34Z","abstract_excerpt":"This paper introduces CryptoBench, the first expert-curated, dynamic benchmark designed to rigorously evaluate the real-world capabilities of Large Language Model (LLM) agents in the uniquely demanding and fast-paced cryptocurrency domain. Unlike general-purpose agent benchmarks for search and prediction, professional crypto analysis presents specific challenges: \\emph{extreme time-sensitivity}, \\emph{a highly adversarial information environment}, and the critical need to synthesize data from \\emph{diverse, specialized sources}, such as on-chain intelligence platforms and real-time Decentraliz"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.00417","kind":"arxiv","version":5},"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/2512.00417/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":"2512.00417","created_at":"2026-05-20T00:01:37.435481+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.00417v5","created_at":"2026-05-20T00:01:37.435481+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.00417","created_at":"2026-05-20T00:01:37.435481+00:00"},{"alias_kind":"pith_short_12","alias_value":"LMDI7SLNZY7V","created_at":"2026-05-20T00:01:37.435481+00:00"},{"alias_kind":"pith_short_16","alias_value":"LMDI7SLNZY7VBQ2N","created_at":"2026-05-20T00:01:37.435481+00:00"},{"alias_kind":"pith_short_8","alias_value":"LMDI7SLN","created_at":"2026-05-20T00:01:37.435481+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.26235","citing_title":"LATTICE: Evaluating Decision Support Utility of Crypto Agents","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD","json":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD.json","graph_json":"https://pith.science/api/pith-number/LMDI7SLNZY7VBQ2NXJW2ANINAD/graph.json","events_json":"https://pith.science/api/pith-number/LMDI7SLNZY7VBQ2NXJW2ANINAD/events.json","paper":"https://pith.science/paper/LMDI7SLN"},"agent_actions":{"view_html":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD","download_json":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD.json","view_paper":"https://pith.science/paper/LMDI7SLN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.00417&json=true","fetch_graph":"https://pith.science/api/pith-number/LMDI7SLNZY7VBQ2NXJW2ANINAD/graph.json","fetch_events":"https://pith.science/api/pith-number/LMDI7SLNZY7VBQ2NXJW2ANINAD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD/action/storage_attestation","attest_author":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD/action/author_attestation","sign_citation":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD/action/citation_signature","submit_replication":"https://pith.science/pith/LMDI7SLNZY7VBQ2NXJW2ANINAD/action/replication_record"}},"created_at":"2026-05-20T00:01:37.435481+00:00","updated_at":"2026-05-20T00:01:37.435481+00:00"}