{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:4T77SKZRVCKP2EVEC76S7HYYAE","short_pith_number":"pith:4T77SKZR","canonical_record":{"source":{"id":"2605.18827","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-12T20:20:06Z","cross_cats_sorted":["cs.LG","cs.PL"],"title_canon_sha256":"2ab67d0d8c79962de8c0449ebedb207d346d8e0bffb37fda5e159eecbbdc3143","abstract_canon_sha256":"7453c72796832d35974c3b093fdfc1ab72922baaddae7eccc5b2cb2a90a73c9f"},"schema_version":"1.0"},"canonical_sha256":"e4fff92b31a894fd12a417fd2f9f18013ea1734b3a424117d03bdc0ada24feb4","source":{"kind":"arxiv","id":"2605.18827","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.18827","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.18827v1","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.18827","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"pith_short_12","alias_value":"4T77SKZRVCKP","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"pith_short_16","alias_value":"4T77SKZRVCKP2EVE","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"pith_short_8","alias_value":"4T77SKZR","created_at":"2026-05-20T00:06:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:4T77SKZRVCKP2EVEC76S7HYYAE","target":"record","payload":{"canonical_record":{"source":{"id":"2605.18827","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-12T20:20:06Z","cross_cats_sorted":["cs.LG","cs.PL"],"title_canon_sha256":"2ab67d0d8c79962de8c0449ebedb207d346d8e0bffb37fda5e159eecbbdc3143","abstract_canon_sha256":"7453c72796832d35974c3b093fdfc1ab72922baaddae7eccc5b2cb2a90a73c9f"},"schema_version":"1.0"},"canonical_sha256":"e4fff92b31a894fd12a417fd2f9f18013ea1734b3a424117d03bdc0ada24feb4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:06:24.676644Z","signature_b64":"lMoV/9dPEOkzq2Y4ye6DQwlCbafZOrmhXFEC3bhDKdZ9zAZd59sVd2TPXp7cchK6MG2QDWnxxg8swAUXSzE7Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4fff92b31a894fd12a417fd2f9f18013ea1734b3a424117d03bdc0ada24feb4","last_reissued_at":"2026-05-20T00:06:24.676053Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:06:24.676053Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.18827","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":"2026-05-20T00:06:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IVIe2PTnYNHgc2dUDhJg7PUMjQuZ+HdiDE0KyR1zIiK2BHkC/JFXpRRv3dtXwPQG8AfPaeqNpz1yHj0Yq93aBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-31T19:21:03.982950Z"},"content_sha256":"bde71c11f43998b5ccd7e37adecbf2516903aa2a4bb653b05cdbbe69b64ef3df","schema_version":"1.0","event_id":"sha256:bde71c11f43998b5ccd7e37adecbf2516903aa2a4bb653b05cdbbe69b64ef3df"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:4T77SKZRVCKP2EVEC76S7HYYAE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Code-Guided Reasoning for Small Language Models: Evaluating Executable MCQA Scaffolds","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","cs.PL"],"primary_cat":"cs.IR","authors_text":"Amit Sheth, Dhaval Patel, Prateek Biswas, Shuxin Lin, Vedant Khandelwal","submitted_at":"2026-05-12T20:20:06Z","abstract_excerpt":"Multiple-choice QA benchmarks usually evaluate small language models (SLMs) as direct answerers, but deployed language-model systems increasingly rely on external scaffolds such as tools, code, and repeated model calls. We introduce Code-Guided Reasoning (CGR), an evaluation protocol and generated-program resource for measuring when executable reasoning scaffolds improve SLM performance on MCQA tasks. CGR standardizes six components: a normalized item interface, a direct solver prompt, a generator prompt, a Python scaffold, solver-call and extraction helpers, and a three-channel result record."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.18827","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/2605.18827/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":"2026-05-20T00:06:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"b50bpGlmeA9PDsJFDKjuoxA84tEjLx+vkgcHD+wSuvWyW8H1bRE9cU1NA5K9X7eufzSFQ5Z5wu+Xe7VU/YTiAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-31T19:21:03.984069Z"},"content_sha256":"04f6da93ec83c02c2ac80fe1d00e7ca5a0baf3ee42758148cee1190335dfa6cf","schema_version":"1.0","event_id":"sha256:04f6da93ec83c02c2ac80fe1d00e7ca5a0baf3ee42758148cee1190335dfa6cf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/4T77SKZRVCKP2EVEC76S7HYYAE/bundle.json","state_url":"https://pith.science/pith/4T77SKZRVCKP2EVEC76S7HYYAE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/4T77SKZRVCKP2EVEC76S7HYYAE/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":"2026-05-31T19:21:03Z","links":{"resolver":"https://pith.science/pith/4T77SKZRVCKP2EVEC76S7HYYAE","bundle":"https://pith.science/pith/4T77SKZRVCKP2EVEC76S7HYYAE/bundle.json","state":"https://pith.science/pith/4T77SKZRVCKP2EVEC76S7HYYAE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/4T77SKZRVCKP2EVEC76S7HYYAE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:4T77SKZRVCKP2EVEC76S7HYYAE","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":"7453c72796832d35974c3b093fdfc1ab72922baaddae7eccc5b2cb2a90a73c9f","cross_cats_sorted":["cs.LG","cs.PL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-12T20:20:06Z","title_canon_sha256":"2ab67d0d8c79962de8c0449ebedb207d346d8e0bffb37fda5e159eecbbdc3143"},"schema_version":"1.0","source":{"id":"2605.18827","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.18827","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.18827v1","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.18827","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"pith_short_12","alias_value":"4T77SKZRVCKP","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"pith_short_16","alias_value":"4T77SKZRVCKP2EVE","created_at":"2026-05-20T00:06:24Z"},{"alias_kind":"pith_short_8","alias_value":"4T77SKZR","created_at":"2026-05-20T00:06:24Z"}],"graph_snapshots":[{"event_id":"sha256:04f6da93ec83c02c2ac80fe1d00e7ca5a0baf3ee42758148cee1190335dfa6cf","target":"graph","created_at":"2026-05-20T00:06:24Z","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/2605.18827/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multiple-choice QA benchmarks usually evaluate small language models (SLMs) as direct answerers, but deployed language-model systems increasingly rely on external scaffolds such as tools, code, and repeated model calls. We introduce Code-Guided Reasoning (CGR), an evaluation protocol and generated-program resource for measuring when executable reasoning scaffolds improve SLM performance on MCQA tasks. CGR standardizes six components: a normalized item interface, a direct solver prompt, a generator prompt, a Python scaffold, solver-call and extraction helpers, and a three-channel result record.","authors_text":"Amit Sheth, Dhaval Patel, Prateek Biswas, Shuxin Lin, Vedant Khandelwal","cross_cats":["cs.LG","cs.PL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-12T20:20:06Z","title":"Code-Guided Reasoning for Small Language Models: Evaluating Executable MCQA Scaffolds"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.18827","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:bde71c11f43998b5ccd7e37adecbf2516903aa2a4bb653b05cdbbe69b64ef3df","target":"record","created_at":"2026-05-20T00:06:24Z","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":"7453c72796832d35974c3b093fdfc1ab72922baaddae7eccc5b2cb2a90a73c9f","cross_cats_sorted":["cs.LG","cs.PL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-12T20:20:06Z","title_canon_sha256":"2ab67d0d8c79962de8c0449ebedb207d346d8e0bffb37fda5e159eecbbdc3143"},"schema_version":"1.0","source":{"id":"2605.18827","kind":"arxiv","version":1}},"canonical_sha256":"e4fff92b31a894fd12a417fd2f9f18013ea1734b3a424117d03bdc0ada24feb4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e4fff92b31a894fd12a417fd2f9f18013ea1734b3a424117d03bdc0ada24feb4","first_computed_at":"2026-05-20T00:06:24.676053Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:06:24.676053Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lMoV/9dPEOkzq2Y4ye6DQwlCbafZOrmhXFEC3bhDKdZ9zAZd59sVd2TPXp7cchK6MG2QDWnxxg8swAUXSzE7Ag==","signature_status":"signed_v1","signed_at":"2026-05-20T00:06:24.676644Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.18827","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bde71c11f43998b5ccd7e37adecbf2516903aa2a4bb653b05cdbbe69b64ef3df","sha256:04f6da93ec83c02c2ac80fe1d00e7ca5a0baf3ee42758148cee1190335dfa6cf"],"state_sha256":"a7ea6d8c412ff937d5cde87f705e4925b289b2117188558e1259ed4cceef3cb7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"splWWVbqMPSgO9XgECozXk9WL6g50TZnRzZMUUmwW6Ekzt+2Bh0pHIh7022CH5PB+eYmX7h0vtyeJQE+SPZeBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-31T19:21:03.988851Z","bundle_sha256":"26d1b437a9e654a1809a09088bf7679531c0c4ce761fef0f5d2f5d88f99c9c64"}}