In-Context Principle Learning from Mistakes Article Swipe
YOU?
·
· 2024
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2402.05403
In-context learning (ICL, also known as few-shot prompting) has been the standard method of adapting LLMs to downstream tasks, by learning from a few input-output examples. Nonetheless, all ICL-based approaches only learn from correct input-output pairs. In this paper, we revisit this paradigm, by learning more from the few given input-output examples. We introduce Learning Principles (LEAP): First, we intentionally induce the model to make mistakes on these few examples; then we reflect on these mistakes, and learn explicit task-specific "principles" from them, which help solve similar problems and avoid common mistakes; finally, we prompt the model to answer unseen test questions using the original few-shot examples and these learned general principles. We evaluate LEAP on a wide range of benchmarks, including multi-hop question answering (Hotpot QA), textual QA (DROP), Big-Bench Hard reasoning, and math problems (GSM8K and MATH); in all these benchmarks, LEAP improves the strongest available LLMs such as GPT-3.5-turbo, GPT-4, GPT-4 turbo and Claude-2.1. For example, LEAP improves over the standard few-shot prompting using GPT-4 by 7.5% in DROP, and by 3.3% in HotpotQA. Importantly, LEAP does not require any more input or examples than the standard few-shot prompting settings.
Related Topics
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2402.05403
- https://arxiv.org/pdf/2402.05403
- OA Status
- green
- Cited By
- 2
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4391709567
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4391709567Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2402.05403Digital Object Identifier
- Title
-
In-Context Principle Learning from MistakesWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2024Year of publication
- Publication date
-
2024-02-08Full publication date if available
- Authors
-
Tianjun Zhang, Aman Madaan, Luyu Gao, Steven Zheng, Swaroop Mishra, Yiming Yang, Niket Tandon, Uri AlonList of authors in order
- Landing page
-
https://arxiv.org/abs/2402.05403Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2402.05403Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2402.05403Direct OA link when available
- Concepts
-
Context (archaeology), Psychology, Cognitive science, Computer science, Epistemology, Geography, Philosophy, ArchaeologyTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
2Total citation count in OpenAlex
- Citations by year (recent)
-
2024: 2Per-year citation counts (last 5 years)
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4391709567 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2402.05403 |
| ids.doi | https://doi.org/10.48550/arxiv.2402.05403 |
| ids.openalex | https://openalex.org/W4391709567 |
| fwci | |
| type | preprint |
| title | In-Context Principle Learning from Mistakes |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T10809 |
| topics[0].field.id | https://openalex.org/fields/36 |
| topics[0].field.display_name | Health Professions |
| topics[0].score | 0.8011000156402588 |
| topics[0].domain.id | https://openalex.org/domains/4 |
| topics[0].domain.display_name | Health Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/3614 |
| topics[0].subfield.display_name | Radiological and Ultrasound Technology |
| topics[0].display_name | Occupational Health and Safety Research |
| topics[1].id | https://openalex.org/T13295 |
| topics[1].field.id | https://openalex.org/fields/22 |
| topics[1].field.display_name | Engineering |
| topics[1].score | 0.7965999841690063 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/2213 |
| topics[1].subfield.display_name | Safety, Risk, Reliability and Quality |
| topics[1].display_name | Safety Systems Engineering in Autonomy |
| topics[2].id | https://openalex.org/T11357 |
| topics[2].field.id | https://openalex.org/fields/18 |
| topics[2].field.display_name | Decision Sciences |
| topics[2].score | 0.7692000269889832 |
| topics[2].domain.id | https://openalex.org/domains/2 |
| topics[2].domain.display_name | Social Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/1804 |
| topics[2].subfield.display_name | Statistics, Probability and Uncertainty |
| topics[2].display_name | Risk and Safety Analysis |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C2779343474 |
| concepts[0].level | 2 |
| concepts[0].score | 0.604851245880127 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q3109175 |
| concepts[0].display_name | Context (archaeology) |
| concepts[1].id | https://openalex.org/C15744967 |
| concepts[1].level | 0 |
| concepts[1].score | 0.4268264174461365 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q9418 |
| concepts[1].display_name | Psychology |
| concepts[2].id | https://openalex.org/C188147891 |
| concepts[2].level | 1 |
| concepts[2].score | 0.38393837213516235 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q147638 |
| concepts[2].display_name | Cognitive science |
| concepts[3].id | https://openalex.org/C41008148 |
| concepts[3].level | 0 |
| concepts[3].score | 0.3508259057998657 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[3].display_name | Computer science |
| concepts[4].id | https://openalex.org/C111472728 |
| concepts[4].level | 1 |
| concepts[4].score | 0.32557186484336853 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q9471 |
| concepts[4].display_name | Epistemology |
| concepts[5].id | https://openalex.org/C205649164 |
| concepts[5].level | 0 |
| concepts[5].score | 0.1824665069580078 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q1071 |
| concepts[5].display_name | Geography |
| concepts[6].id | https://openalex.org/C138885662 |
| concepts[6].level | 0 |
| concepts[6].score | 0.1502012312412262 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q5891 |
| concepts[6].display_name | Philosophy |
| concepts[7].id | https://openalex.org/C166957645 |
| concepts[7].level | 1 |
| concepts[7].score | 0.0 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q23498 |
| concepts[7].display_name | Archaeology |
| keywords[0].id | https://openalex.org/keywords/context |
| keywords[0].score | 0.604851245880127 |
| keywords[0].display_name | Context (archaeology) |
| keywords[1].id | https://openalex.org/keywords/psychology |
| keywords[1].score | 0.4268264174461365 |
| keywords[1].display_name | Psychology |
| keywords[2].id | https://openalex.org/keywords/cognitive-science |
| keywords[2].score | 0.38393837213516235 |
| keywords[2].display_name | Cognitive science |
| keywords[3].id | https://openalex.org/keywords/computer-science |
| keywords[3].score | 0.3508259057998657 |
| keywords[3].display_name | Computer science |
| keywords[4].id | https://openalex.org/keywords/epistemology |
| keywords[4].score | 0.32557186484336853 |
| keywords[4].display_name | Epistemology |
| keywords[5].id | https://openalex.org/keywords/geography |
| keywords[5].score | 0.1824665069580078 |
| keywords[5].display_name | Geography |
| keywords[6].id | https://openalex.org/keywords/philosophy |
| keywords[6].score | 0.1502012312412262 |
| keywords[6].display_name | Philosophy |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2402.05403 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400194 |
| locations[0].source.issn | |
| locations[0].source.type | repository |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | arXiv (Cornell University) |
| locations[0].source.host_organization | https://openalex.org/I205783295 |
| locations[0].source.host_organization_name | Cornell University |
| locations[0].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[0].license | |
| locations[0].pdf_url | https://arxiv.org/pdf/2402.05403 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2402.05403 |
| locations[1].id | doi:10.48550/arxiv.2402.05403 |
| locations[1].is_oa | True |
| locations[1].source.id | https://openalex.org/S4306400194 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | True |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | arXiv (Cornell University) |
| locations[1].source.host_organization | https://openalex.org/I205783295 |
| locations[1].source.host_organization_name | Cornell University |
| locations[1].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[1].license | cc-by |
| locations[1].pdf_url | |
| locations[1].version | |
| locations[1].raw_type | article |
| locations[1].license_id | https://openalex.org/licenses/cc-by |
| locations[1].is_accepted | False |
| locations[1].is_published | |
| locations[1].raw_source_name | |
| locations[1].landing_page_url | https://doi.org/10.48550/arxiv.2402.05403 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5101571114 |
| authorships[0].author.orcid | https://orcid.org/0000-0003-2059-0709 |
| authorships[0].author.display_name | Tianjun Zhang |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Zhang, Tianjun |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5091998799 |
| authorships[1].author.orcid | |
| authorships[1].author.display_name | Aman Madaan |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Madaan, Aman |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5060885692 |
| authorships[2].author.orcid | https://orcid.org/0009-0006-5806-3022 |
| authorships[2].author.display_name | Luyu Gao |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Gao, Luyu |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5112919416 |
| authorships[3].author.orcid | |
| authorships[3].author.display_name | Steven Zheng |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Zheng, Steven |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5063722751 |
| authorships[4].author.orcid | https://orcid.org/0009-0001-6413-7001 |
| authorships[4].author.display_name | Swaroop Mishra |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | Mishra, Swaroop |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5041434096 |
| authorships[5].author.orcid | https://orcid.org/0000-0003-1359-0364 |
| authorships[5].author.display_name | Yiming Yang |
| authorships[5].author_position | middle |
| authorships[5].raw_author_name | Yang, Yiming |
| authorships[5].is_corresponding | False |
| authorships[6].author.id | https://openalex.org/A5052882677 |
| authorships[6].author.orcid | |
| authorships[6].author.display_name | Niket Tandon |
| authorships[6].author_position | middle |
| authorships[6].raw_author_name | Tandon, Niket |
| authorships[6].is_corresponding | False |
| authorships[7].author.id | https://openalex.org/A5107243336 |
| authorships[7].author.orcid | https://orcid.org/0000-0001-5796-9988 |
| authorships[7].author.display_name | Uri Alon |
| authorships[7].author_position | last |
| authorships[7].raw_author_name | Alon, Uri |
| authorships[7].is_corresponding | False |
| has_content.pdf | False |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2402.05403 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-10T00:00:00 |
| display_name | In-Context Principle Learning from Mistakes |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T10809 |
| primary_topic.field.id | https://openalex.org/fields/36 |
| primary_topic.field.display_name | Health Professions |
| primary_topic.score | 0.8011000156402588 |
| primary_topic.domain.id | https://openalex.org/domains/4 |
| primary_topic.domain.display_name | Health Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/3614 |
| primary_topic.subfield.display_name | Radiological and Ultrasound Technology |
| primary_topic.display_name | Occupational Health and Safety Research |
| related_works | https://openalex.org/W2748952813, https://openalex.org/W2390279801, https://openalex.org/W2358668433, https://openalex.org/W2376932109, https://openalex.org/W2001405890, https://openalex.org/W2382290278, https://openalex.org/W2478288626, https://openalex.org/W2350741829, https://openalex.org/W2530322880, https://openalex.org/W1596801655 |
| cited_by_count | 2 |
| counts_by_year[0].year | 2024 |
| counts_by_year[0].cited_by_count | 2 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2402.05403 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400194 |
| best_oa_location.source.issn | |
| best_oa_location.source.type | repository |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | arXiv (Cornell University) |
| best_oa_location.source.host_organization | https://openalex.org/I205783295 |
| best_oa_location.source.host_organization_name | Cornell University |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| best_oa_location.license | |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2402.05403 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | http://arxiv.org/abs/2402.05403 |
| primary_location.id | pmh:oai:arXiv.org:2402.05403 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400194 |
| primary_location.source.issn | |
| primary_location.source.type | repository |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | arXiv (Cornell University) |
| primary_location.source.host_organization | https://openalex.org/I205783295 |
| primary_location.source.host_organization_name | Cornell University |
| primary_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| primary_location.license | |
| primary_location.pdf_url | https://arxiv.org/pdf/2402.05403 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2402.05403 |
| publication_date | 2024-02-08 |
| publication_year | 2024 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 22, 116 |
| abstract_inverted_index.In | 36 |
| abstract_inverted_index.QA | 128 |
| abstract_inverted_index.We | 52, 112 |
| abstract_inverted_index.as | 5, 150 |
| abstract_inverted_index.by | 19, 43, 168, 173 |
| abstract_inverted_index.in | 139, 170, 175 |
| abstract_inverted_index.of | 13, 119 |
| abstract_inverted_index.on | 66, 73, 115 |
| abstract_inverted_index.or | 185 |
| abstract_inverted_index.to | 16, 63, 97 |
| abstract_inverted_index.we | 39, 58, 71, 93 |
| abstract_inverted_index.For | 157 |
| abstract_inverted_index.all | 27, 140 |
| abstract_inverted_index.and | 76, 88, 107, 133, 137, 155, 172 |
| abstract_inverted_index.any | 182 |
| abstract_inverted_index.few | 23, 48, 68 |
| abstract_inverted_index.has | 8 |
| abstract_inverted_index.not | 180 |
| abstract_inverted_index.the | 10, 47, 61, 95, 103, 145, 162, 188 |
| abstract_inverted_index.3.3% | 174 |
| abstract_inverted_index.7.5% | 169 |
| abstract_inverted_index.Hard | 131 |
| abstract_inverted_index.LEAP | 114, 143, 159, 178 |
| abstract_inverted_index.LLMs | 15, 148 |
| abstract_inverted_index.QA), | 126 |
| abstract_inverted_index.also | 3 |
| abstract_inverted_index.been | 9 |
| abstract_inverted_index.does | 179 |
| abstract_inverted_index.from | 21, 32, 46, 81 |
| abstract_inverted_index.help | 84 |
| abstract_inverted_index.make | 64 |
| abstract_inverted_index.math | 134 |
| abstract_inverted_index.more | 45, 183 |
| abstract_inverted_index.only | 30 |
| abstract_inverted_index.over | 161 |
| abstract_inverted_index.such | 149 |
| abstract_inverted_index.test | 100 |
| abstract_inverted_index.than | 187 |
| abstract_inverted_index.then | 70 |
| abstract_inverted_index.this | 37, 41 |
| abstract_inverted_index.wide | 117 |
| abstract_inverted_index.(ICL, | 2 |
| abstract_inverted_index.DROP, | 171 |
| abstract_inverted_index.GPT-4 | 153, 167 |
| abstract_inverted_index.avoid | 89 |
| abstract_inverted_index.given | 49 |
| abstract_inverted_index.input | 184 |
| abstract_inverted_index.known | 4 |
| abstract_inverted_index.learn | 31, 77 |
| abstract_inverted_index.model | 62, 96 |
| abstract_inverted_index.range | 118 |
| abstract_inverted_index.solve | 85 |
| abstract_inverted_index.them, | 82 |
| abstract_inverted_index.these | 67, 74, 108, 141 |
| abstract_inverted_index.turbo | 154 |
| abstract_inverted_index.using | 102, 166 |
| abstract_inverted_index.which | 83 |
| abstract_inverted_index.(GSM8K | 136 |
| abstract_inverted_index.First, | 57 |
| abstract_inverted_index.GPT-4, | 152 |
| abstract_inverted_index.MATH); | 138 |
| abstract_inverted_index.answer | 98 |
| abstract_inverted_index.common | 90 |
| abstract_inverted_index.induce | 60 |
| abstract_inverted_index.method | 12 |
| abstract_inverted_index.pairs. | 35 |
| abstract_inverted_index.paper, | 38 |
| abstract_inverted_index.prompt | 94 |
| abstract_inverted_index.tasks, | 18 |
| abstract_inverted_index.unseen | 99 |
| abstract_inverted_index.(DROP), | 129 |
| abstract_inverted_index.(Hotpot | 125 |
| abstract_inverted_index.(LEAP): | 56 |
| abstract_inverted_index.correct | 33 |
| abstract_inverted_index.general | 110 |
| abstract_inverted_index.learned | 109 |
| abstract_inverted_index.reflect | 72 |
| abstract_inverted_index.require | 181 |
| abstract_inverted_index.revisit | 40 |
| abstract_inverted_index.similar | 86 |
| abstract_inverted_index.textual | 127 |
| abstract_inverted_index.Learning | 54 |
| abstract_inverted_index.adapting | 14 |
| abstract_inverted_index.evaluate | 113 |
| abstract_inverted_index.example, | 158 |
| abstract_inverted_index.examples | 106, 186 |
| abstract_inverted_index.explicit | 78 |
| abstract_inverted_index.few-shot | 6, 105, 164, 190 |
| abstract_inverted_index.finally, | 92 |
| abstract_inverted_index.improves | 144, 160 |
| abstract_inverted_index.learning | 1, 20, 44 |
| abstract_inverted_index.mistakes | 65 |
| abstract_inverted_index.original | 104 |
| abstract_inverted_index.problems | 87, 135 |
| abstract_inverted_index.question | 123 |
| abstract_inverted_index.standard | 11, 163, 189 |
| abstract_inverted_index.Big-Bench | 130 |
| abstract_inverted_index.HotpotQA. | 176 |
| abstract_inverted_index.ICL-based | 28 |
| abstract_inverted_index.answering | 124 |
| abstract_inverted_index.available | 147 |
| abstract_inverted_index.examples. | 25, 51 |
| abstract_inverted_index.examples; | 69 |
| abstract_inverted_index.including | 121 |
| abstract_inverted_index.introduce | 53 |
| abstract_inverted_index.mistakes, | 75 |
| abstract_inverted_index.mistakes; | 91 |
| abstract_inverted_index.multi-hop | 122 |
| abstract_inverted_index.paradigm, | 42 |
| abstract_inverted_index.prompting | 165, 191 |
| abstract_inverted_index.questions | 101 |
| abstract_inverted_index.settings. | 192 |
| abstract_inverted_index.strongest | 146 |
| abstract_inverted_index.In-context | 0 |
| abstract_inverted_index.Principles | 55 |
| abstract_inverted_index.approaches | 29 |
| abstract_inverted_index.downstream | 17 |
| abstract_inverted_index.prompting) | 7 |
| abstract_inverted_index.reasoning, | 132 |
| abstract_inverted_index.Claude-2.1. | 156 |
| abstract_inverted_index.benchmarks, | 120, 142 |
| abstract_inverted_index.principles. | 111 |
| abstract_inverted_index."principles" | 80 |
| abstract_inverted_index.Importantly, | 177 |
| abstract_inverted_index.Nonetheless, | 26 |
| abstract_inverted_index.input-output | 24, 34, 50 |
| abstract_inverted_index.intentionally | 59 |
| abstract_inverted_index.task-specific | 79 |
| abstract_inverted_index.GPT-3.5-turbo, | 151 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 8 |
| citation_normalized_percentile |