Produced as part of the SERI ML Alignment Theory Scholars Program with support from Cavendish Labs, Many thanks go to the following for reading and commenting: Dennis Akar, Walter Laurito, Soroush Pour, Kaarel Hänni

Goals of post:

Awareness of current inductive bias research related to human-like decisions
Understand the known mechanisms behind inductive bias development
Compare similarities of humans and why this could be important
What’s the bigger picture importance and potential alignment research connection

Aim to answer these questions:

What does human cognition-inspired inductive bias look like in a model?
Is this divide in model and human bias simply a result of the different types of data that humans take in vs the model?
What are the mechanics behind certain inductive biases in humans?
Is it possible that these findings could translate to a generalization of other human-like decisions in AI?
What elements help create the conditions necessary to shift a model’s inductive bias to match that of a human?

(Future post: 2. Potential ways to create further convergence, experimental ideas)

What is inductive bias

Models

In machine learning, when referring to the inductive bias of a particular architecture and training process, we are pointing to the distribution of models it produces. For example, if a CNN is trained to classify a training set of red apples vs green pears, is it more likely to become a red-vs-green-object-classifier, or a round-vs-pointy-object-classifier (or something entirely different). The term ‘bias’ in this case refers to any basis for choosing one generalization over another one, differing from strictly following the observed training instances.

Inductive bias can be thought of as the set of rules governing what kind of algorithm is likely to end up being implemented in the model architecture, including patterns in how it will generalize to test data and out-of-distribution. This allows the model to generalize. Occam's razor is a classic example of inductive bias, which says the simplest consistent explanation is generally the most likely to be the correct one.

There are some known factors that contribute to the creation of a model’s inductive bias, such as: choice in model architecture, training data, feature selection/preprocessing, and optimization algorithms such as Stochastic Gradient Descent (SGD) can all introduce ...

Posts

Wikitag Contributions

Comments