{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:X7AKAICZVVODKPTBJ2XEXKUWUY","short_pith_number":"pith:X7AKAICZ","canonical_record":{"source":{"id":"2409.07638","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-09-11T21:48:33Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"4069a514c24064cfeb642bd347c5e4be1ed8478e7b6c289e79c3fa107d814ac0","abstract_canon_sha256":"f80b3dd49d6153844a8ff5cae37d189d86c60e819665a1e8b731af30fa3b76c1"},"schema_version":"1.0"},"canonical_sha256":"bfc0a02059ad5c353e614eae4baa96a61673447eaad28df35e695ce9f1b00ac3","source":{"kind":"arxiv","id":"2409.07638","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.07638","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"arxiv_version","alias_value":"2409.07638v2","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.07638","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"pith_short_12","alias_value":"X7AKAICZVVOD","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"pith_short_16","alias_value":"X7AKAICZVVODKPTB","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"pith_short_8","alias_value":"X7AKAICZ","created_at":"2026-07-05T09:11:16Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:X7AKAICZVVODKPTBJ2XEXKUWUY","target":"record","payload":{"canonical_record":{"source":{"id":"2409.07638","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-09-11T21:48:33Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"4069a514c24064cfeb642bd347c5e4be1ed8478e7b6c289e79c3fa107d814ac0","abstract_canon_sha256":"f80b3dd49d6153844a8ff5cae37d189d86c60e819665a1e8b731af30fa3b76c1"},"schema_version":"1.0"},"canonical_sha256":"bfc0a02059ad5c353e614eae4baa96a61673447eaad28df35e695ce9f1b00ac3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:11:16.838527Z","signature_b64":"1j0+g83nQY5o1gs4ROpTgYhYpVFq9u9xOGHQcB/HR7QM86Lg+U7iyAP9uPnShdufL9z7MueRvOr1feeY6r0fCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bfc0a02059ad5c353e614eae4baa96a61673447eaad28df35e695ce9f1b00ac3","last_reissued_at":"2026-07-05T09:11:16.838068Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:11:16.838068Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.07638","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:11:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kNTru2tad//0kr6jaoizZ8ahxJa4UlfrGzU0pRHFUuEL/1qYdGkOX5wx81x/QaIQyRdGVUoPQzTx+KM2VNEjAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T19:01:23.762700Z"},"content_sha256":"a78baf14d9cf1c59f5a7f595786a802ac79b927d5ffdbee2c401f96844622ad6","schema_version":"1.0","event_id":"sha256:a78baf14d9cf1c59f5a7f595786a802ac79b927d5ffdbee2c401f96844622ad6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:X7AKAICZVVODKPTBJ2XEXKUWUY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can We Count on LLMs? The Fixed-Effect Fallacy and Claims of GPT-4 Capabilities","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Cormac Herley, Shuo Chen, Thomas Ball","submitted_at":"2024-09-11T21:48:33Z","abstract_excerpt":"In this paper we explore evaluation of LLM capabilities. We present measurements of GPT-4 performance on several deterministic tasks; each task involves a basic calculation and takes as input parameter some element drawn from a large well-defined population (e.g., count elements in a list, multiply two k-digit numbers, etc). We examine several conditions per-task and perform enough trials so that statistically significant differences can be detected. This allows us to investigate the sensitivity of task-accuracy both to query phrasing and input parameter population. We find that seemingly triv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.07638","kind":"arxiv","version":2},"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/2409.07638/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":"2026-07-05T09:11:16Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vwjZpwda4d7KISPa2jVtzPiF/l8cCudDXOceGPzekj6mL/fl3ohTvR05DrMwRlPs0kPOjV4mn8//E4FxVNXOAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T19:01:23.763227Z"},"content_sha256":"c5aa3a84c41fea355658e324679adcf4895d97ffac7d359bd3b610bdcfbbecc9","schema_version":"1.0","event_id":"sha256:c5aa3a84c41fea355658e324679adcf4895d97ffac7d359bd3b610bdcfbbecc9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/X7AKAICZVVODKPTBJ2XEXKUWUY/bundle.json","state_url":"https://pith.science/pith/X7AKAICZVVODKPTBJ2XEXKUWUY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/X7AKAICZVVODKPTBJ2XEXKUWUY/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":"2026-08-16T19:01:23Z","links":{"resolver":"https://pith.science/pith/X7AKAICZVVODKPTBJ2XEXKUWUY","bundle":"https://pith.science/pith/X7AKAICZVVODKPTBJ2XEXKUWUY/bundle.json","state":"https://pith.science/pith/X7AKAICZVVODKPTBJ2XEXKUWUY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/X7AKAICZVVODKPTBJ2XEXKUWUY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:X7AKAICZVVODKPTBJ2XEXKUWUY","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":"f80b3dd49d6153844a8ff5cae37d189d86c60e819665a1e8b731af30fa3b76c1","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-09-11T21:48:33Z","title_canon_sha256":"4069a514c24064cfeb642bd347c5e4be1ed8478e7b6c289e79c3fa107d814ac0"},"schema_version":"1.0","source":{"id":"2409.07638","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.07638","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"arxiv_version","alias_value":"2409.07638v2","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.07638","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"pith_short_12","alias_value":"X7AKAICZVVOD","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"pith_short_16","alias_value":"X7AKAICZVVODKPTB","created_at":"2026-07-05T09:11:16Z"},{"alias_kind":"pith_short_8","alias_value":"X7AKAICZ","created_at":"2026-07-05T09:11:16Z"}],"graph_snapshots":[{"event_id":"sha256:c5aa3a84c41fea355658e324679adcf4895d97ffac7d359bd3b610bdcfbbecc9","target":"graph","created_at":"2026-07-05T09:11:16Z","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/2409.07638/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In this paper we explore evaluation of LLM capabilities. We present measurements of GPT-4 performance on several deterministic tasks; each task involves a basic calculation and takes as input parameter some element drawn from a large well-defined population (e.g., count elements in a list, multiply two k-digit numbers, etc). We examine several conditions per-task and perform enough trials so that statistically significant differences can be detected. This allows us to investigate the sensitivity of task-accuracy both to query phrasing and input parameter population. We find that seemingly triv","authors_text":"Cormac Herley, Shuo Chen, Thomas Ball","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-09-11T21:48:33Z","title":"Can We Count on LLMs? The Fixed-Effect Fallacy and Claims of GPT-4 Capabilities"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.07638","kind":"arxiv","version":2},"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:a78baf14d9cf1c59f5a7f595786a802ac79b927d5ffdbee2c401f96844622ad6","target":"record","created_at":"2026-07-05T09:11:16Z","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":"f80b3dd49d6153844a8ff5cae37d189d86c60e819665a1e8b731af30fa3b76c1","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-09-11T21:48:33Z","title_canon_sha256":"4069a514c24064cfeb642bd347c5e4be1ed8478e7b6c289e79c3fa107d814ac0"},"schema_version":"1.0","source":{"id":"2409.07638","kind":"arxiv","version":2}},"canonical_sha256":"bfc0a02059ad5c353e614eae4baa96a61673447eaad28df35e695ce9f1b00ac3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bfc0a02059ad5c353e614eae4baa96a61673447eaad28df35e695ce9f1b00ac3","first_computed_at":"2026-07-05T09:11:16.838068Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:11:16.838068Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1j0+g83nQY5o1gs4ROpTgYhYpVFq9u9xOGHQcB/HR7QM86Lg+U7iyAP9uPnShdufL9z7MueRvOr1feeY6r0fCg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:11:16.838527Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.07638","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a78baf14d9cf1c59f5a7f595786a802ac79b927d5ffdbee2c401f96844622ad6","sha256:c5aa3a84c41fea355658e324679adcf4895d97ffac7d359bd3b610bdcfbbecc9"],"state_sha256":"2c50f41fe70f0a1f4d5372274fbe6bb9ce21f348142849f9762e19d95aa5c0ae"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"O239LAwOWlyx00q2IGu82OlyuGJQpBofN9AnqycD8QYYO3uHxi9D+DDom4/qR35o8l94jnX3MvSBz1xv3+i7BA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-16T19:01:23.766922Z","bundle_sha256":"0b644f68225d18dd501bf0ab778a22991f91c2ed2428bd5ab9eb97287bc780f6"}}