Irina Rish

papers

462

total citations

papers (11)

Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization

NEURIPS 2021arXiv

325

citations

Context is Key: A Benchmark for Forecasting with Essential Textual Information

ICML 2025arXiv

citations

Parametric Scattering Networks

CVPR 2022arXiv

citations

Maximum State Entropy Exploration using Predecessor and Successor Representations

NEURIPS 2023arXiv

citations

Seq-VCR: Preventing Collapse in Intermediate Transformer Representations for Enhanced Reasoning

ICLR 2025arXiv

citations

Online Fast Adaptation and Knowledge Accumulation (OSAKA): a New Approach to Continual Learning

NEURIPS 2020

citations

Adversarial Feature Desensitization

NEURIPS 2021arXiv

citations

Irina Rish

papers (11)

Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization

Context is Key: A Benchmark for Forecasting with Essential Textual Information

Parametric Scattering Networks

Maximum State Entropy Exploration using Predecessor and Successor Representations

Continual Learning In Environments With Polynomial Mixing Times

Surprising Effectiveness of pretraining Ternary Language Model at Scale

Unsupervised Concept Discovery Mitigates Spurious Correlations

Non-Adversarial Inverse Reinforcement Learning via Successor Feature Matching

Seq-VCR: Preventing Collapse in Intermediate Transformer Representations for Enhanced Reasoning

Online Fast Adaptation and Knowledge Accumulation (OSAKA): a New Approach to Continual Learning

Adversarial Feature Desensitization

papers (11)

Invariance Principle Meets Information Bottleneck for Out-of-Distribution Generalization

Context is Key: A Benchmark for Forecasting with Essential Textual Information

Parametric Scattering Networks

Maximum State Entropy Exploration using Predecessor and Successor Representations

Continual Learning In Environments With Polynomial Mixing Times

Surprising Effectiveness of pretraining Ternary Language Model at Scale

Unsupervised Concept Discovery Mitigates Spurious Correlations

Non-Adversarial Inverse Reinforcement Learning via Successor Feature Matching

Seq-VCR: Preventing Collapse in Intermediate Transformer Representations for Enhanced Reasoning

Online Fast Adaptation and Knowledge Accumulation (OSAKA): a New Approach to Continual Learning

Adversarial Feature Desensitization