{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:PVQFLF7HT33P5AJKR3LBFHQUDB","short_pith_number":"pith:PVQFLF7H","canonical_record":{"source":{"id":"2507.03834","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-04T23:16:02Z","cross_cats_sorted":[],"title_canon_sha256":"9da67b77d08cff91861fdf1720e1aed007c1aa1bb7509069901c07dada10ac47","abstract_canon_sha256":"37b82c9438988a5f6132adb1c7137092018bf697b2e4c6050bf9aa7effa950ca"},"schema_version":"1.0"},"canonical_sha256":"7d605597e79ef6fe812a8ed6129e14187712eb74570379841fcf8cba54c44058","source":{"kind":"arxiv","id":"2507.03834","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.03834","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"arxiv_version","alias_value":"2507.03834v1","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.03834","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"pith_short_12","alias_value":"PVQFLF7HT33P","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"pith_short_16","alias_value":"PVQFLF7HT33P5AJK","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"pith_short_8","alias_value":"PVQFLF7H","created_at":"2026-07-05T11:32:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:PVQFLF7HT33P5AJKR3LBFHQUDB","target":"record","payload":{"canonical_record":{"source":{"id":"2507.03834","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-04T23:16:02Z","cross_cats_sorted":[],"title_canon_sha256":"9da67b77d08cff91861fdf1720e1aed007c1aa1bb7509069901c07dada10ac47","abstract_canon_sha256":"37b82c9438988a5f6132adb1c7137092018bf697b2e4c6050bf9aa7effa950ca"},"schema_version":"1.0"},"canonical_sha256":"7d605597e79ef6fe812a8ed6129e14187712eb74570379841fcf8cba54c44058","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:32:35.171277Z","signature_b64":"qNJ/HIEsEapIVehbXx9aeIDZMm5YYeEEZtzIUnPYrEej+/Gomj38EBr4aqckfafDOYoLIxz2z8FOWO5iuODCDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d605597e79ef6fe812a8ed6129e14187712eb74570379841fcf8cba54c44058","last_reissued_at":"2026-07-05T11:32:35.170830Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:32:35.170830Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2507.03834","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:32:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PgDtBU97JIE9N//uwm39hZ2f8ablgjx1VOMGsxivW7hCrTmPdI98oUqc2DMCLKKIcSmb9pXa0gBnQ6kaldq0AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T10:38:01.345580Z"},"content_sha256":"ae7c6c26c4c23d91e47013dae8b57c7d25229e13672f85ca1909b18d285ea345","schema_version":"1.0","event_id":"sha256:ae7c6c26c4c23d91e47013dae8b57c7d25229e13672f85ca1909b18d285ea345"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:PVQFLF7HT33P5AJKR3LBFHQUDB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Economic Evaluation of LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Matt Thomson, Michael J. Zellinger","submitted_at":"2025-07-04T23:16:02Z","abstract_excerpt":"Practitioners often navigate LLM performance trade-offs by plotting Pareto frontiers of optimal accuracy-cost trade-offs. However, this approach offers no way to compare between LLMs with distinct strengths and weaknesses: for example, a cheap, error-prone model vs a pricey but accurate one. To address this gap, we propose economic evaluation of LLMs. Our framework quantifies the performance trade-off of an LLM as a single number based on the economic constraints of a concrete use case, all expressed in dollars: the cost of making a mistake, the cost of incremental latency, and the cost of abs"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.03834","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/2507.03834/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:32:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"c7biE8PnyQMfNNsveribKsO92kg9uyazU2/2si4D2ll7lEctcJOg+6I9naBOl5fv4pvY+hm0oUq6sbyTcvhICA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T10:38:01.346107Z"},"content_sha256":"76e30076c8dee670ef802c461243ec97ecceda658f35e96c44bf528de9bff0cd","schema_version":"1.0","event_id":"sha256:76e30076c8dee670ef802c461243ec97ecceda658f35e96c44bf528de9bff0cd"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/PVQFLF7HT33P5AJKR3LBFHQUDB/bundle.json","state_url":"https://pith.science/pith/PVQFLF7HT33P5AJKR3LBFHQUDB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/PVQFLF7HT33P5AJKR3LBFHQUDB/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-07T10:38:01Z","links":{"resolver":"https://pith.science/pith/PVQFLF7HT33P5AJKR3LBFHQUDB","bundle":"https://pith.science/pith/PVQFLF7HT33P5AJKR3LBFHQUDB/bundle.json","state":"https://pith.science/pith/PVQFLF7HT33P5AJKR3LBFHQUDB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/PVQFLF7HT33P5AJKR3LBFHQUDB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:PVQFLF7HT33P5AJKR3LBFHQUDB","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"37b82c9438988a5f6132adb1c7137092018bf697b2e4c6050bf9aa7effa950ca","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-04T23:16:02Z","title_canon_sha256":"9da67b77d08cff91861fdf1720e1aed007c1aa1bb7509069901c07dada10ac47"},"schema_version":"1.0","source":{"id":"2507.03834","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.03834","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"arxiv_version","alias_value":"2507.03834v1","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.03834","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"pith_short_12","alias_value":"PVQFLF7HT33P","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"pith_short_16","alias_value":"PVQFLF7HT33P5AJK","created_at":"2026-07-05T11:32:35Z"},{"alias_kind":"pith_short_8","alias_value":"PVQFLF7H","created_at":"2026-07-05T11:32:35Z"}],"graph_snapshots":[{"event_id":"sha256:76e30076c8dee670ef802c461243ec97ecceda658f35e96c44bf528de9bff0cd","target":"graph","created_at":"2026-07-05T11:32:35Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2507.03834/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Practitioners often navigate LLM performance trade-offs by plotting Pareto frontiers of optimal accuracy-cost trade-offs. However, this approach offers no way to compare between LLMs with distinct strengths and weaknesses: for example, a cheap, error-prone model vs a pricey but accurate one. To address this gap, we propose economic evaluation of LLMs. Our framework quantifies the performance trade-off of an LLM as a single number based on the economic constraints of a concrete use case, all expressed in dollars: the cost of making a mistake, the cost of incremental latency, and the cost of abs","authors_text":"Matt Thomson, Michael J. Zellinger","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-04T23:16:02Z","title":"Economic Evaluation of LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.03834","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ae7c6c26c4c23d91e47013dae8b57c7d25229e13672f85ca1909b18d285ea345","target":"record","created_at":"2026-07-05T11:32:35Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"37b82c9438988a5f6132adb1c7137092018bf697b2e4c6050bf9aa7effa950ca","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-04T23:16:02Z","title_canon_sha256":"9da67b77d08cff91861fdf1720e1aed007c1aa1bb7509069901c07dada10ac47"},"schema_version":"1.0","source":{"id":"2507.03834","kind":"arxiv","version":1}},"canonical_sha256":"7d605597e79ef6fe812a8ed6129e14187712eb74570379841fcf8cba54c44058","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7d605597e79ef6fe812a8ed6129e14187712eb74570379841fcf8cba54c44058","first_computed_at":"2026-07-05T11:32:35.170830Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:32:35.170830Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qNJ/HIEsEapIVehbXx9aeIDZMm5YYeEEZtzIUnPYrEej+/Gomj38EBr4aqckfafDOYoLIxz2z8FOWO5iuODCDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:32:35.171277Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.03834","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ae7c6c26c4c23d91e47013dae8b57c7d25229e13672f85ca1909b18d285ea345","sha256:76e30076c8dee670ef802c461243ec97ecceda658f35e96c44bf528de9bff0cd"],"state_sha256":"8aac87504aee7796ddf7639c002972861ecbb51027c752daf5d39414181b088e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zVzi0+uSQ1k82aD/GA4zIjadLSOfIYB1UStDgEfhUHnKvoB0Vm/R++zyDRVHPWDYD0eoNBsAKCOVWxxJ+k7LAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T10:38:01.351995Z","bundle_sha256":"a126c838a8e9163d992e1a4b1c9ebc913fc0d684d326cbd948cb2f78cad07d56"}}