{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZMIFYRSOBFPPOGLPP4UE7YPKM5","short_pith_number":"pith:ZMIFYRSO","schema_version":"1.0","canonical_sha256":"cb105c464e095ef7196f7f284fe1ea674c3184fac406bf5ee338be79779cfea8","source":{"kind":"arxiv","id":"2503.17354","version":1},"attestation_state":"computed","paper":{"title":"HCAST: Human-Calibrated Autonomy Software Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Amy Deng, Aron Lajko, Ben West, Brian Goodrich, Chris Canal, Daniel O'Connel, David Rein, Elizabeth Barnes, Joel Becker, Katharyn Garcia, Lawrence Chan, Lucas Jun Koba Sato, Max Hasin, Megan Kinniment, Nate Rush, Pip Arnott, Ryan Bloom, Sami Jawhar, Seraphina Nix, Sydney von Arx, Thomas Broadley, Thomas Kwa","submitted_at":"2025-03-21T17:54:01Z","abstract_excerpt":"To understand and predict the societal impacts of highly autonomous AI systems, we need benchmarks with grounding, i.e., metrics that directly connect AI performance to real-world effects we care about. We present HCAST (Human-Calibrated Autonomy Software Tasks), a benchmark of 189 machine learning engineering, cybersecurity, software engineering, and general reasoning tasks. We collect 563 human baselines (totaling over 1500 hours) from people skilled in these domains, working under identical conditions as AI agents, which lets us estimate that HCAST tasks take humans between one minute and 8"},"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":"2503.17354","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-03-21T17:54:01Z","cross_cats_sorted":[],"title_canon_sha256":"172c742a06c83d4c650c4f629b6523a538d00d5028e7c0a805cc040b6c6f2419","abstract_canon_sha256":"b55ce11d56ac02256f84abea88f4652455964796a70c70f0bf19b587a0c18ccf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:37:04.245492Z","signature_b64":"cv8a0NllrOqYe8EJLfJJVwtEpSILDMqBzgivUb61hfp37E+VD0GoEn92d5H81MdiZPkAbqwDonYRj1l0B3v5DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cb105c464e095ef7196f7f284fe1ea674c3184fac406bf5ee338be79779cfea8","last_reissued_at":"2026-07-05T10:37:04.244783Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:37:04.244783Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HCAST: Human-Calibrated Autonomy Software Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Amy Deng, Aron Lajko, Ben West, Brian Goodrich, Chris Canal, Daniel O'Connel, David Rein, Elizabeth Barnes, Joel Becker, Katharyn Garcia, Lawrence Chan, Lucas Jun Koba Sato, Max Hasin, Megan Kinniment, Nate Rush, Pip Arnott, Ryan Bloom, Sami Jawhar, Seraphina Nix, Sydney von Arx, Thomas Broadley, Thomas Kwa","submitted_at":"2025-03-21T17:54:01Z","abstract_excerpt":"To understand and predict the societal impacts of highly autonomous AI systems, we need benchmarks with grounding, i.e., metrics that directly connect AI performance to real-world effects we care about. We present HCAST (Human-Calibrated Autonomy Software Tasks), a benchmark of 189 machine learning engineering, cybersecurity, software engineering, and general reasoning tasks. We collect 563 human baselines (totaling over 1500 hours) from people skilled in these domains, working under identical conditions as AI agents, which lets us estimate that HCAST tasks take humans between one minute and 8"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.17354","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/2503.17354/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":"2503.17354","created_at":"2026-07-05T10:37:04.244871+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.17354v1","created_at":"2026-07-05T10:37:04.244871+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.17354","created_at":"2026-07-05T10:37:04.244871+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZMIFYRSOBFPP","created_at":"2026-07-05T10:37:04.244871+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZMIFYRSOBFPPOGLP","created_at":"2026-07-05T10:37:04.244871+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZMIFYRSO","created_at":"2026-07-05T10:37:04.244871+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07157","citing_title":"Think Fast: Estimating No-CoT Task-Completion Time Horizons of Frontier AI Models","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00913","citing_title":"Two AI Metrics Diverged: Will it Make All the Difference?","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07157","citing_title":"Think Fast: Estimating No-CoT Task-Completion Time Horizons of Frontier AI Models","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11926","citing_title":"Toward Generalist Autonomous Research via Hypothesis-Tree Refinement","ref_index":146,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05912","citing_title":"FrontierFinance: A Long-Horizon Computer-Use Benchmark of Real-World Financial Tasks","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5","json":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5.json","graph_json":"https://pith.science/api/pith-number/ZMIFYRSOBFPPOGLPP4UE7YPKM5/graph.json","events_json":"https://pith.science/api/pith-number/ZMIFYRSOBFPPOGLPP4UE7YPKM5/events.json","paper":"https://pith.science/paper/ZMIFYRSO"},"agent_actions":{"view_html":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5","download_json":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5.json","view_paper":"https://pith.science/paper/ZMIFYRSO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.17354&json=true","fetch_graph":"https://pith.science/api/pith-number/ZMIFYRSOBFPPOGLPP4UE7YPKM5/graph.json","fetch_events":"https://pith.science/api/pith-number/ZMIFYRSOBFPPOGLPP4UE7YPKM5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5/action/storage_attestation","attest_author":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5/action/author_attestation","sign_citation":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5/action/citation_signature","submit_replication":"https://pith.science/pith/ZMIFYRSOBFPPOGLPP4UE7YPKM5/action/replication_record"}},"created_at":"2026-07-05T10:37:04.244871+00:00","updated_at":"2026-07-05T10:37:04.244871+00:00"}