{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:S3APCICNMPAEJ6EPPEIVRIED3W","short_pith_number":"pith:S3APCICN","schema_version":"1.0","canonical_sha256":"96c0f1204d63c044f88f791158a083ddaaad8fe8f2b28fb026162ae26d2504ff","source":{"kind":"arxiv","id":"2312.04333","version":4},"attestation_state":"computed","paper":{"title":"Is Bigger and Deeper Always Better? Probing LLaMA Across Scales and Layers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dongmei Zhang, Jia Li, Linjun Shou, Ming Gong, Ning Wu, Nuo Chen, Shining Liang","submitted_at":"2023-12-07T14:50:41Z","abstract_excerpt":"This paper presents an in-depth analysis of Large Language Models (LLMs), focusing on LLaMA, a prominent open-source foundational model in natural language processing. Instead of assessing LLaMA through its generative output, we design multiple-choice tasks to probe its intrinsic understanding in high-order tasks such as reasoning and computation. We examine the model horizontally, comparing different sizes, and vertically, assessing different layers. We unveil several key and uncommon findings based on the designed probing tasks: (1) Horizontally, enlarging model sizes almost could not automa"},"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":"2312.04333","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-07T14:50:41Z","cross_cats_sorted":[],"title_canon_sha256":"6d47b7757ac21e598fe69c64a61a155a6f72911b29f2166e05812d53ead526f2","abstract_canon_sha256":"5151bb04ee97a254d89ae5f62e3dd030e02c8b25b572be5a6b9d91256457c70c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:31:35.583615Z","signature_b64":"IKf9uja7skSc58kWLSei9x7bIYN937w6NkN4zZSrGWMMyDfQT7p88rVw2LPl1Q7SJ7zptwTJd4pcp8zA5VWPDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"96c0f1204d63c044f88f791158a083ddaaad8fe8f2b28fb026162ae26d2504ff","last_reissued_at":"2026-07-05T07:31:35.583090Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:31:35.583090Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Is Bigger and Deeper Always Better? Probing LLaMA Across Scales and Layers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dongmei Zhang, Jia Li, Linjun Shou, Ming Gong, Ning Wu, Nuo Chen, Shining Liang","submitted_at":"2023-12-07T14:50:41Z","abstract_excerpt":"This paper presents an in-depth analysis of Large Language Models (LLMs), focusing on LLaMA, a prominent open-source foundational model in natural language processing. Instead of assessing LLaMA through its generative output, we design multiple-choice tasks to probe its intrinsic understanding in high-order tasks such as reasoning and computation. We examine the model horizontally, comparing different sizes, and vertically, assessing different layers. We unveil several key and uncommon findings based on the designed probing tasks: (1) Horizontally, enlarging model sizes almost could not automa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.04333","kind":"arxiv","version":4},"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/2312.04333/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":"2312.04333","created_at":"2026-07-05T07:31:35.583150+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.04333v4","created_at":"2026-07-05T07:31:35.583150+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.04333","created_at":"2026-07-05T07:31:35.583150+00:00"},{"alias_kind":"pith_short_12","alias_value":"S3APCICNMPAE","created_at":"2026-07-05T07:31:35.583150+00:00"},{"alias_kind":"pith_short_16","alias_value":"S3APCICNMPAEJ6EP","created_at":"2026-07-05T07:31:35.583150+00:00"},{"alias_kind":"pith_short_8","alias_value":"S3APCICN","created_at":"2026-07-05T07:31:35.583150+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.22567","citing_title":"LANG: Reinforcement Learning for Multilingual Reasoning with Language-Adaptive Hint Guidance","ref_index":57,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W","json":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W.json","graph_json":"https://pith.science/api/pith-number/S3APCICNMPAEJ6EPPEIVRIED3W/graph.json","events_json":"https://pith.science/api/pith-number/S3APCICNMPAEJ6EPPEIVRIED3W/events.json","paper":"https://pith.science/paper/S3APCICN"},"agent_actions":{"view_html":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W","download_json":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W.json","view_paper":"https://pith.science/paper/S3APCICN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.04333&json=true","fetch_graph":"https://pith.science/api/pith-number/S3APCICNMPAEJ6EPPEIVRIED3W/graph.json","fetch_events":"https://pith.science/api/pith-number/S3APCICNMPAEJ6EPPEIVRIED3W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W/action/storage_attestation","attest_author":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W/action/author_attestation","sign_citation":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W/action/citation_signature","submit_replication":"https://pith.science/pith/S3APCICNMPAEJ6EPPEIVRIED3W/action/replication_record"}},"created_at":"2026-07-05T07:31:35.583150+00:00","updated_at":"2026-07-05T07:31:35.583150+00:00"}