paper
referenced-only
2019
paper:arxiv-1907-02893Invariant Risk Minimization
ByMartín Arjovsky·L. Bottou·Ishaan Gulrajani·David Lopez-Paz
Original abstract (expand)
We introduce Invariant Risk Minimization (IRM), a learning paradigm to estimate invariant correlations across multiple training distributions. To achieve this goal, IRM learns a data representation such that the optimal classifier, on top of that data representation, matches for all training distributions. Through theory and experiments, we show how the invariances learned by IRM relate to the causal structures governing the data and enable out-of-distribution generalization.
Similar preprints — Semantic Scholar
Cited by (1)
- Understanding and Steering Llama 3 with Sparse Autoencoders - Goodfire
A sparse autoencoder trained on layer-19 residual-stream activations of Llama-3-8B-Instruct, using LMSYS-Chat-1M, yields features that can be labeled automatically, ranked by causal relevance, and use