y0news
AnalyticsDigestsSourcesTopicsRSSAICrypto

#robust-ai News & Analysis

2 articles tagged with #robust-ai. AI-curated summaries with sentiment analysis and key takeaways from 50+ sources.

2 articles
AINeutralarXiv – CS AI · Mar 47/103
🧠

Unsupervised Representation Learning -- an Invariant Risk Minimization Perspective

Researchers propose a new unsupervised framework for Invariant Risk Minimization (IRM) that learns robust representations without labeled data. The approach introduces two methods - Principal Invariant Component Analysis (PICA) and Variational Invariant Autoencoder (VIAE) - that can capture invariant structures across different environments using only unlabeled data.

AINeutralarXiv – CS AI · Mar 174/10
🧠

Learning When to Trust in Contextual Bandits

Researchers propose CESA-LinUCB, a new approach to robust reinforcement learning that addresses 'Contextual Sycophancy' where evaluators are truthful in normal situations but biased in critical contexts. The method learns trust boundaries for each evaluator and achieves sublinear regret even when no evaluator is globally reliable.