Open-Sourcing Highly Capable Foundation Models: An evaluation of risks, benefits, and alternative methods for pursuing open-source objectives.
Elizabeth SegerNoemi DrekslerRichard MoulangeEmily DardamanJonas SchuettK. WeiChristoph WinterMackenzie ArnoldSeán Ó hÉigeartaighAnton KorinekMarkus AnderljungBen BucknallAlan ChanEoghan StaffordLeonie KoesslerAviv OvadyaBen GarfinkelEmma BluemkeMichael AirdPatrick LevermoreJulian HazellAbhishek GuptaPublished in: CoRR (2023)
Keyphrases
- alternative methods
- open source
- case study
- probabilistic model
- statistical models
- evaluation method
- multi objective
- parameter estimation
- computational models
- neural network model
- experimental data
- business process
- regression model
- machine learning algorithms
- data sets
- model selection
- source code
- decision trees
- neural network