Login / Signup

Probing BERT's priors with serial reproduction chains.

Takateru YamakoshiThomas L. GriffithsRobert D. Hawkins
Published in: ACL (Findings) (2022)
Keyphrases
  • prior information
  • bayesian framework
  • heavy tailed
  • lower bound
  • special case
  • prior probabilities
  • learned from training data
  • neural network
  • training data
  • maximum likelihood
  • model selection
  • prior distribution