Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms.
Rafael RafailovYaswanth ChittepuRyan ParkHarshit SikchiJoey HejnaW. Bradley KnoxChelsea FinnScott NiekumPublished in: CoRR (2024)
Keyphrases
- probabilistic model
- theoretical analysis
- linear models
- computational model
- statistical model
- mathematical model
- high level
- formal model
- computationally efficient
- optimization problems
- computational cost
- prior knowledge
- computational complexity
- management system
- markov chain
- data structure
- objective function
- experimental data
- decision problems
- statistical methods
- similarity measure
- neural network