Score breakdown

Textbooks Are All You Need

paper-0172 · paper · 2023

Suriya Gunasekar et al. (Microsoft Research)

Primary technical report for a notable AI model (verified primary source).

Abstract

We introduce phi-1, a new large language model for code, with significantly smaller size than competing models: phi-1 is a Transformer-based model with 1.3B parameters, trained for 4 days on 8 A100s, using a selection of ``textbook quality" data from the web (6B tokens) and synthetically generated textbooks and exercises with GPT-3.5 (1B tokens). Despite this small scale, phi-1 attains pass@1 accuracy 50.6% on HumanEval and 55.5% on MBPP. It also displays surprising emergent properties compared to phi-1-base, our model before our finetuning stage on a dataset of coding exercises, and phi-1-small, a smaller model with 350M parameters trained with the same pipeline as phi-1 that still achieves 45% on HumanEval. [OpenAlex]

Academic, score -0.2105

MetricStatusValueNorm.WeightContributionSourceConfidenceLicenseProvenance
citation_countpresent99.00.0004410.50.000221OpenAlexhighOpenAlex, CC0 metadatalink
library_holdingsmissingrecorded as missing, penalized by rule, never imputed−0.1recorded as missing; penalized by rule, never imputed
readership_persistencepresent5.00.2857140.050.014286OpenAlexmediumOpenAlex, CC0 metadatalink
syllabus_adoptionsmissingrecorded as missing, penalized by rule, never imputed−0.125recorded as missing; penalized by rule, never imputed

Broad Influence, score -0.0856

MetricStatusValueNorm.WeightContributionSourceConfidenceLicenseProvenance
citation_countpresent99.00.0004410.28.8e-05OpenAlexhighOpenAlex, CC0 metadatalink
library_holdingsmissingrecorded as missing, penalized by rule, never imputed−0.125recorded as missing; penalized by rule, never imputed
readership_persistencepresent5.00.2857140.40.114286OpenAlexmediumOpenAlex, CC0 metadatalink
syllabus_adoptionsmissingrecorded as missing, penalized by rule, never imputed−0.075recorded as missing; penalized by rule, never imputed

Governance Practitioner, score -0.2963

MetricStatusValueNorm.WeightContributionSourceConfidenceLicenseProvenance
citation_countpresent99.00.0004410.250.00011OpenAlexhighOpenAlex, CC0 metadatalink
library_holdingsmissingrecorded as missing, penalized by rule, never imputed−0.15recorded as missing; penalized by rule, never imputed
readership_persistencepresent5.00.2857140.10.028571OpenAlexmediumOpenAlex, CC0 metadatalink
syllabus_adoptionsmissingrecorded as missing, penalized by rule, never imputed−0.175recorded as missing; penalized by rule, never imputed

A rank is not a verdict on intrinsic worth. It is a transparent output of declared evidence, weights, and missing-data rules at a specific release date.

Disagree with this rank or a number? Challenge it with your evidence. Every challenge gets a public identifier and a published resolution.