Does Invariant Risk Minimization Capture Invariance?

Pritish Kamath · Akilesh Tangella · Danica Sutherland · Nathan Srebro


Keywords: [ Models and Methods ] [ Multi-task and transfer learning ]

[ Abstract ]
[ Slides
Tue 13 Apr 2 p.m. PDT — 4 p.m. PDT
Oral presentation: Generalization / Reinforcement Learning / Optimization
Thu 15 Apr 1 p.m. PDT — 2 p.m. PDT


We show that the Invariant Risk Minimization (IRM) formulation of Arjovsky et al. (2019) can fail to capture "natural" invariances, at least when used in its practical "linear" form, and even on very simple problems which directly follow the motivating examples for IRM. This can lead to worse generalization on new environments, even when compared to unconstrained ERM. The issue stems from a significant gap between the linear variant (as in their concrete method IRMv1) and the full non-linear IRM formulation. Additionally, even when capturing the "right" invariances, we show that it is possible for IRM to learn a sub-optimal predictor, due to the loss function not being invariant across environments. The issues arise even when measuring invariance on the population distributions, but are exacerbated by the fact that IRM is extremely fragile to sampling.

Chat is not available.