{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:IST3XHVNJSADRSBL5L6JRTVLNT","short_pith_number":"pith:IST3XHVN","schema_version":"1.0","canonical_sha256":"44a7bb9ead4c8038c82beafc98ceab6ce87f1de79c8e0b9e50615d9cc1fee7af","source":{"kind":"arxiv","id":"2304.03277","version":1},"attestation_state":"computed","paper":{"title":"Instruction Tuning with GPT-4","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"GPT-4 generated instruction data enables LLaMA models to reach higher zero-shot performance on new tasks than earlier synthetic datasets.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Baolin Peng, Chunyuan Li, Jianfeng Gao, Michel Galley, Pengcheng He","submitted_at":"2023-04-06T17:58:09Z","abstract_excerpt":"Prior work has shown that finetuning large language models (LLMs) using machine-generated instruction-following data enables such models to achieve remarkable zero-shot capabilities on new tasks, and no human-written instructions are needed. In this paper, we present the first attempt to use GPT-4 to generate instruction-following data for LLM finetuning. Our early experiments on instruction-tuned LLaMA models show that the 52K English and Chinese instruction-following data generated by GPT-4 leads to superior zero-shot performance on new tasks to the instruction-following data generated by pr"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2304.03277","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-04-06T17:58:09Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2da0c9d0dd1bb0ca8ff88374f47294a26d976e55871dbc5d18aa2d9f2cae193e","abstract_canon_sha256":"a5fa7e08c14cfe870725a2b0ab026cc82993bc6daed0af6062dcbe35be363a91"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:22.360701Z","signature_b64":"S+UED9yYs0+zmAMcnpwQloKfVlHgac9qhW074tDP0h0a3aGJj/IFy9dMI9WPbE3j1ApZ4w4ww/V7Q2v3N0DzAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"44a7bb9ead4c8038c82beafc98ceab6ce87f1de79c8e0b9e50615d9cc1fee7af","last_reissued_at":"2026-05-17T23:39:22.359932Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:22.359932Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Instruction Tuning with GPT-4","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"GPT-4 generated instruction data enables LLaMA models to reach higher zero-shot performance on new tasks than earlier synthetic datasets.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Baolin Peng, Chunyuan Li, Jianfeng Gao, Michel Galley, Pengcheng He","submitted_at":"2023-04-06T17:58:09Z","abstract_excerpt":"Prior work has shown that finetuning large language models (LLMs) using machine-generated instruction-following data enables such models to achieve remarkable zero-shot capabilities on new tasks, and no human-written instructions are needed. In this paper, we present the first attempt to use GPT-4 to generate instruction-following data for LLM finetuning. Our early experiments on instruction-tuned LLaMA models show that the 52K English and Chinese instruction-following data generated by GPT-4 leads to superior zero-shot performance on new tasks to the instruction-following data generated by pr"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our early experiments on instruction-tuned LLaMA models show that the 52K English and Chinese instruction-following data generated by GPT-4 leads to superior zero-shot performance on new tasks to the instruction-following data generated by previous state-of-the-art models.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That performance gains on the chosen zero-shot tasks reflect genuine improvement in instruction following rather than artifacts of GPT-4's data generation process or evaluation choices.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"GPT-4-generated instruction data produces superior zero-shot performance in finetuned LLaMA models versus prior state-of-the-art data.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"GPT-4 generated instruction data enables LLaMA models to reach higher zero-shot performance on new tasks than earlier synthetic datasets.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"3e5864af4ec74bde1b4fb3b6168781c91bd2d762b48e55b99d23d6306483114d"},"source":{"id":"2304.03277","kind":"arxiv","version":1},"verdict":{"id":"7170c55c-42cb-42e8-b30e-0c41f8928e3d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T16:57:59.957977Z","strongest_claim":"Our early experiments on instruction-tuned LLaMA models show that the 52K English and Chinese instruction-following data generated by GPT-4 leads to superior zero-shot performance on new tasks to the instruction-following data generated by previous state-of-the-art models.","one_line_summary":"GPT-4-generated instruction data produces superior zero-shot performance in finetuned LLaMA models versus prior state-of-the-art data.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That performance gains on the chosen zero-shot tasks reflect genuine improvement in instruction following rather than artifacts of GPT-4's data generation process or evaluation choices.","pith_extraction_headline":"GPT-4 generated instruction data enables LLaMA models to reach higher zero-shot performance on new tasks than earlier synthetic datasets."},"references":{"count":18,"sample":[{"doi":"","year":null,"title":"A General Language Assistant as a Laboratory for Alignment","work_id":"a43f9ea0-01be-47d5-b8ee-a1a9f73381c5","ref_index":1,"cited_arxiv_id":"2112.00861","is_internal_anchor":true},{"doi":"10.5281/zenodo.7733589","year":null,"title":"Yejin Bang, Samuel Cahyawijaya, Nayeon Lee, Wenliang Dai, Dan Su, Bryan Wilie, Holy Lovenia, Ziwei Ji, Tiezheng Yu, Willy Chung, et al","work_id":"6f218053-cca5-4fde-92aa-730589931f0c","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Constitutional AI: Harmlessness from AI Feedback","work_id":"faaaa4e0-2676-4fac-a0b4-99aef10d2095","ref_index":3,"cited_arxiv_id":"2212.08073","is_internal_anchor":true},{"doi":"10.5281/zenodo.5297715","year":1901,"title":"GPT-Neo: Large Scale Autoregressive Language Modeling with Mesh-Tensorflow , March 2021","work_id":"6c7f8a44-6f52-448c-b819-5ba82a7bbc59","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Scaling Instruction-Finetuned Language Models","work_id":"8405abb1-7558-4fdf-af24-f4c52fa77a06","ref_index":5,"cited_arxiv_id":"2210.11416","is_internal_anchor":true}],"resolved_work":18,"snapshot_sha256":"086a76bdf9860cd1a85670949cbde7831e84f7f71ab1d678f23db21772a9c2ed","internal_anchors":11},"formal_canon":{"evidence_count":1,"snapshot_sha256":"c26f16a8dcc7d1eb015e86884555556a48859d117881c5f57bf53f828be0fea0"},"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":"2304.03277","created_at":"2026-05-17T23:39:22.360054+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.03277v1","created_at":"2026-05-17T23:39:22.360054+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.03277","created_at":"2026-05-17T23:39:22.360054+00:00"},{"alias_kind":"pith_short_12","alias_value":"IST3XHVNJSAD","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"IST3XHVNJSADRSBL","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"IST3XHVN","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":42,"internal_anchor_count":42,"sample":[{"citing_arxiv_id":"2310.13289","citing_title":"SALMONN: Towards Generic Hearing Abilities for Large Language Models","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2309.05653","citing_title":"MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuning","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2307.04964","citing_title":"Secrets of RLHF in Large Language Models Part I: PPO","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2305.17926","citing_title":"Large Language Models are not Fair Evaluators","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2305.07895","citing_title":"OCRBench: On the Hidden Mystery of OCR in Large Multimodal Models","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2311.07575","citing_title":"SPHINX: The Joint Mixing of Weights, Tasks, and Visual Embeddings for Multi-modal Large Language Models","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2310.14566","citing_title":"HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination and Visual Illusion in Large Vision-Language Models","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2512.13751","citing_title":"MIDUS: Memory-Infused Depth Up-Scaling","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2311.01378","citing_title":"Vision-Language Foundation Models as Effective Robot Imitators","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2601.14249","citing_title":"Which Reasoning Trajectories Teach Students to Reason Better? A Simple Metric of Informative Alignment","ref_index":31,"is_internal_anchor":true},{"citing_arxiv_id":"2306.13549","citing_title":"A Survey on Multimodal Large Language Models","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2402.09353","citing_title":"DoRA: Weight-Decomposed Low-Rank Adaptation","ref_index":96,"is_internal_anchor":true},{"citing_arxiv_id":"2410.07283","citing_title":"Prompt Infection: LLM-to-LLM Prompt Injection within Multi-Agent Systems","ref_index":77,"is_internal_anchor":true},{"citing_arxiv_id":"2410.07283","citing_title":"Prompt Infection: LLM-to-LLM Prompt Injection within Multi-Agent Systems","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2305.18323","citing_title":"ReWOO: Decoupling Reasoning from Observations for Efficient Augmented Language Models","ref_index":37,"is_internal_anchor":true},{"citing_arxiv_id":"2410.04417","citing_title":"SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference","ref_index":61,"is_internal_anchor":true},{"citing_arxiv_id":"2304.15010","citing_title":"LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model","ref_index":50,"is_internal_anchor":true},{"citing_arxiv_id":"2305.03726","citing_title":"Otter: A Multi-Modal Model with In-Context Instruction Tuning","ref_index":69,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14530","citing_title":"Mitigating Mask Prior Drift and Positional Attention Collapse in Large Diffusion Vision-Language Models","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2303.16199","citing_title":"LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention","ref_index":64,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13328","citing_title":"What Limits Vision-and-Language Navigation ?","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2509.20328","citing_title":"Video models are zero-shot learners and reasoners","ref_index":78,"is_internal_anchor":true},{"citing_arxiv_id":"2401.10020","citing_title":"Self-Rewarding Language Models","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11922","citing_title":"StepCodeReasoner: Aligning Code Reasoning with Stepwise Execution Traces via Reinforcement Learning","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11376","citing_title":"LLM-X: A Scalable Negotiation-Oriented Exchange for Communication Among Personal LLM Agents","ref_index":24,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":1,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT","json":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT.json","graph_json":"https://pith.science/api/pith-number/IST3XHVNJSADRSBL5L6JRTVLNT/graph.json","events_json":"https://pith.science/api/pith-number/IST3XHVNJSADRSBL5L6JRTVLNT/events.json","paper":"https://pith.science/paper/2304.03277"},"agent_actions":{"view_html":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT","download_json":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT.json","view_paper":"https://pith.science/paper/2304.03277","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.03277&json=true","fetch_graph":"https://pith.science/api/pith-number/IST3XHVNJSADRSBL5L6JRTVLNT/graph.json","fetch_events":"https://pith.science/api/pith-number/IST3XHVNJSADRSBL5L6JRTVLNT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT/action/storage_attestation","attest_author":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT/action/author_attestation","sign_citation":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT/action/citation_signature","submit_replication":"https://pith.science/pith/IST3XHVNJSADRSBL5L6JRTVLNT/action/replication_record"}},"created_at":"2026-05-17T23:39:22.360054+00:00","updated_at":"2026-05-17T23:39:22.360054+00:00"}