{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:7JQWOXS7EHYAUC2HPHL2QFGC25","short_pith_number":"pith:7JQWOXS7","schema_version":"1.0","canonical_sha256":"fa61675e5f21f00a0b4779d7a814c2d77c4c9d3c46ed93701fe97e537462a18e","source":{"kind":"arxiv","id":"2101.06804","version":1},"attestation_state":"computed","paper":{"title":"What Makes Good In-Context Examples for GPT-$3$?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bill Dolan, Dinghan Shen, Jiachang Liu, Lawrence Carin, Weizhu Chen, Yizhe Zhang","submitted_at":"2021-01-17T23:38:40Z","abstract_excerpt":"GPT-$3$ has attracted lots of attention due to its superior performance across a wide range of NLP tasks, especially with its powerful and versatile in-context few-shot learning ability. Despite its success, we found that the empirical results of GPT-$3$ depend heavily on the choice of in-context examples. In this work, we investigate whether there are more effective strategies for judiciously selecting in-context examples (relative to random sampling) that better leverage GPT-$3$'s few-shot capabilities. Inspired by the recent success of leveraging a retrieval module to augment large-scale ne"},"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":"2101.06804","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-01-17T23:38:40Z","cross_cats_sorted":[],"title_canon_sha256":"10f50c41dc032bdbc8a28b15be0070950cfe6611cff39b54a5e7d3cd677d54bf","abstract_canon_sha256":"f5ecb3a71b0949e73f9a42009dba6d2127b54e2da9b44d0d1a1da408d5a62ddc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:07:37.281067Z","signature_b64":"KBXdujPKyn/SBfAy4bEDsFN4jMc7ccLG+o7tNxuBd1day9FdUMSVrsyppEn/LgSfwPoXseuwjMKrtgELmLpHDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fa61675e5f21f00a0b4779d7a814c2d77c4c9d3c46ed93701fe97e537462a18e","last_reissued_at":"2026-07-05T02:07:37.280622Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:07:37.280622Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What Makes Good In-Context Examples for GPT-$3$?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bill Dolan, Dinghan Shen, Jiachang Liu, Lawrence Carin, Weizhu Chen, Yizhe Zhang","submitted_at":"2021-01-17T23:38:40Z","abstract_excerpt":"GPT-$3$ has attracted lots of attention due to its superior performance across a wide range of NLP tasks, especially with its powerful and versatile in-context few-shot learning ability. Despite its success, we found that the empirical results of GPT-$3$ depend heavily on the choice of in-context examples. In this work, we investigate whether there are more effective strategies for judiciously selecting in-context examples (relative to random sampling) that better leverage GPT-$3$'s few-shot capabilities. Inspired by the recent success of leveraging a retrieval module to augment large-scale ne"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2101.06804","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/2101.06804/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":"2101.06804","created_at":"2026-07-05T02:07:37.280692+00:00"},{"alias_kind":"arxiv_version","alias_value":"2101.06804v1","created_at":"2026-07-05T02:07:37.280692+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2101.06804","created_at":"2026-07-05T02:07:37.280692+00:00"},{"alias_kind":"pith_short_12","alias_value":"7JQWOXS7EHYA","created_at":"2026-07-05T02:07:37.280692+00:00"},{"alias_kind":"pith_short_16","alias_value":"7JQWOXS7EHYAUC2H","created_at":"2026-07-05T02:07:37.280692+00:00"},{"alias_kind":"pith_short_8","alias_value":"7JQWOXS7","created_at":"2026-07-05T02:07:37.280692+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":26,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06452","citing_title":"From Voting to Agent Collaboration: Answer-Type-Aware LLM Pipelines for BioASQ 14b","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2606.24937","citing_title":"The Hitchhiker's Guide to Agentic AI: From Foundations to Systems","ref_index":130,"is_internal_anchor":false},{"citing_arxiv_id":"2606.18620","citing_title":"BCL: Bayesian In-Context Learning Framework for Information Extraction","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02423","citing_title":"Neuron-Aware Active Few-Shot Learning for LLMs","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08068","citing_title":"DICE: Entropy-Regularized Equilibrium Selection for Stable Multi-Agent LLM Coordination","ref_index":81,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30775","citing_title":"A Single Rewrite Suffices: Empirical Lessons from Production Skill Description Optimization","ref_index":80,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24718","citing_title":"The Tokenizer Tax Across 25 European Languages: Domain Invariance, Cross-Lingual Few-Shot Effects, and the Ukrainian Penalty","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2502.16022","citing_title":"Enhancing LLMs for Identifying and Prioritizing Important Medical Jargons from Electronic Health Record Notes Utilizing Data Augmentation: A Comparative Study","ref_index":96,"is_internal_anchor":false},{"citing_arxiv_id":"2504.02181","citing_title":"A Survey of Scaling in Large Language Model Reasoning","ref_index":113,"is_internal_anchor":false},{"citing_arxiv_id":"2504.20605","citing_title":"TF1-EN-3M: Three Million Synthetic Moral Fables for Training Small, Open Language Models","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18635","citing_title":"Data Presentation Over Architecture: Resampling Strategies for Credit Risk Prediction with Tabular Foundation Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2307.06435","citing_title":"A Comprehensive Overview of Large Language Models","ref_index":207,"is_internal_anchor":false},{"citing_arxiv_id":"2509.11295","citing_title":"The Prompt Engineering Report Distilled: Quick Start Guide for Life Sciences","ref_index":82,"is_internal_anchor":false},{"citing_arxiv_id":"2509.11295","citing_title":"The Prompt Engineering Report Distilled: Quick Start Guide for Life Sciences","ref_index":83,"is_internal_anchor":false},{"citing_arxiv_id":"2104.08773","citing_title":"Cross-Task Generalization via Natural Language Crowdsourcing Instructions","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2511.02135","citing_title":"Graph-Based Alternatives to LLMs for Human Simulation","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2301.12652","citing_title":"REPLUG: Retrieval-Augmented Black-Box Language Models","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2202.12837","citing_title":"Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?","ref_index":216,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04174","citing_title":"CoALFake: Collaborative Active Learning with Human-LLM Co-Annotation for Cross-Domain Fake News Detection","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2205.11916","citing_title":"Large Language Models are Zero-Shot Reasoners","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2204.14198","citing_title":"Flamingo: a Visual Language Model for Few-Shot Learning","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2305.05176","citing_title":"FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2304.03442","citing_title":"Generative Agents: Interactive Simulacra of Human Behavior","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21018","citing_title":"Adaptive Test-Time Compute Allocation with Evolving In-Context Demonstrations","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06627","citing_title":"DiffuMask: Diffusion Language Model for Token-level Prompt Pruning","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25","json":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25.json","graph_json":"https://pith.science/api/pith-number/7JQWOXS7EHYAUC2HPHL2QFGC25/graph.json","events_json":"https://pith.science/api/pith-number/7JQWOXS7EHYAUC2HPHL2QFGC25/events.json","paper":"https://pith.science/paper/7JQWOXS7"},"agent_actions":{"view_html":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25","download_json":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25.json","view_paper":"https://pith.science/paper/7JQWOXS7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2101.06804&json=true","fetch_graph":"https://pith.science/api/pith-number/7JQWOXS7EHYAUC2HPHL2QFGC25/graph.json","fetch_events":"https://pith.science/api/pith-number/7JQWOXS7EHYAUC2HPHL2QFGC25/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25/action/storage_attestation","attest_author":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25/action/author_attestation","sign_citation":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25/action/citation_signature","submit_replication":"https://pith.science/pith/7JQWOXS7EHYAUC2HPHL2QFGC25/action/replication_record"}},"created_at":"2026-07-05T02:07:37.280692+00:00","updated_at":"2026-07-05T02:07:37.280692+00:00"}