{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:Z4BDUHSKR2SOD3NZQAE6QWRGDW","short_pith_number":"pith:Z4BDUHSK","schema_version":"1.0","canonical_sha256":"cf023a1e4a8ea4e1edb98009e85a261da083e3e7ecb5fa0f49679d3f8c5ec85b","source":{"kind":"arxiv","id":"2306.14050","version":2},"attestation_state":"computed","paper":{"title":"Symbolic Chain-of-Thought Distillation: Small Models Can Also \"Think\" Step-by-Step","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jack Hessel, Kai-Wei Chang, Liunian Harold Li, Xiang Ren, Yejin Choi, Youngjae Yu","submitted_at":"2023-06-24T20:15:07Z","abstract_excerpt":"Chain-of-thought prompting (e.g., \"Let's think step-by-step\") primes large language models to verbalize rationalization for their predictions. While chain-of-thought can lead to dramatic performance gains, benefits appear to emerge only for sufficiently large models (beyond 50B parameters). We show that orders-of-magnitude smaller models (125M -- 1.3B parameters) can still benefit from chain-of-thought prompting. To achieve this, we introduce Symbolic Chain-of-Thought Distillation (SCoTD), a method to train a smaller student model on rationalizations sampled from a significantly larger teacher"},"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":"2306.14050","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-06-24T20:15:07Z","cross_cats_sorted":[],"title_canon_sha256":"18d582a2efbabe177b974ac1e8fa03a10857b440295d2a6b44b9e6ea58158864","abstract_canon_sha256":"1781b6482417a6c5df2239e197ea745bdcf0f4b89add46cd4e1c02cd9f065af9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:08:17.255365Z","signature_b64":"UtGsyOCbNytzhNn4bq3TQxb9hBNRY4i2Yhw4u7ew9CKEg3YHQlTcfwrIPFHyFzjtT+W2JlCs8vfg5rm1ivBFBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cf023a1e4a8ea4e1edb98009e85a261da083e3e7ecb5fa0f49679d3f8c5ec85b","last_reissued_at":"2026-07-05T08:08:17.254938Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:08:17.254938Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Symbolic Chain-of-Thought Distillation: Small Models Can Also \"Think\" Step-by-Step","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jack Hessel, Kai-Wei Chang, Liunian Harold Li, Xiang Ren, Yejin Choi, Youngjae Yu","submitted_at":"2023-06-24T20:15:07Z","abstract_excerpt":"Chain-of-thought prompting (e.g., \"Let's think step-by-step\") primes large language models to verbalize rationalization for their predictions. While chain-of-thought can lead to dramatic performance gains, benefits appear to emerge only for sufficiently large models (beyond 50B parameters). We show that orders-of-magnitude smaller models (125M -- 1.3B parameters) can still benefit from chain-of-thought prompting. To achieve this, we introduce Symbolic Chain-of-Thought Distillation (SCoTD), a method to train a smaller student model on rationalizations sampled from a significantly larger teacher"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.14050","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/2306.14050/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":"2306.14050","created_at":"2026-07-05T08:08:17.254993+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.14050v2","created_at":"2026-07-05T08:08:17.254993+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.14050","created_at":"2026-07-05T08:08:17.254993+00:00"},{"alias_kind":"pith_short_12","alias_value":"Z4BDUHSKR2SO","created_at":"2026-07-05T08:08:17.254993+00:00"},{"alias_kind":"pith_short_16","alias_value":"Z4BDUHSKR2SOD3NZ","created_at":"2026-07-05T08:08:17.254993+00:00"},{"alias_kind":"pith_short_8","alias_value":"Z4BDUHSK","created_at":"2026-07-05T08:08:17.254993+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2305.02301","citing_title":"Distilling Step-by-Step! Outperforming Larger Language Models with Less Training Data and Smaller Model Sizes","ref_index":84,"is_internal_anchor":false},{"citing_arxiv_id":"2510.25977","citing_title":"NeuronMLP: Efficient LLM Inference via Singular Value Decomposition Compression and Tiling on AWS Trainium","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2401.05459","citing_title":"Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Security","ref_index":259,"is_internal_anchor":false},{"citing_arxiv_id":"2404.14294","citing_title":"A Survey on Efficient Inference for Large Language Models","ref_index":121,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07307","citing_title":"Rethinking Dense Sequential Chains: Reasoning Language Models Can Extract Answers from Sparse, Order-Shuffling Chain-of-Thoughts","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW","json":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW.json","graph_json":"https://pith.science/api/pith-number/Z4BDUHSKR2SOD3NZQAE6QWRGDW/graph.json","events_json":"https://pith.science/api/pith-number/Z4BDUHSKR2SOD3NZQAE6QWRGDW/events.json","paper":"https://pith.science/paper/Z4BDUHSK"},"agent_actions":{"view_html":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW","download_json":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW.json","view_paper":"https://pith.science/paper/Z4BDUHSK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.14050&json=true","fetch_graph":"https://pith.science/api/pith-number/Z4BDUHSKR2SOD3NZQAE6QWRGDW/graph.json","fetch_events":"https://pith.science/api/pith-number/Z4BDUHSKR2SOD3NZQAE6QWRGDW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW/action/storage_attestation","attest_author":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW/action/author_attestation","sign_citation":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW/action/citation_signature","submit_replication":"https://pith.science/pith/Z4BDUHSKR2SOD3NZQAE6QWRGDW/action/replication_record"}},"created_at":"2026-07-05T08:08:17.254993+00:00","updated_at":"2026-07-05T08:08:17.254993+00:00"}