#causal-learning News & Analysis

4 articles tagged with #causal-learning. AI-curated summaries with sentiment analysis and key takeaways from 50+ sources.

4 articles

AIBullisharXiv – CS AI · Mar 167/10

🧠

Active Causal Structure Learning with Latent Variables: Towards Learning to Detour in Autonomous Robots

Researchers propose Active Causal Structure Learning with Latent Variables (ACSLWL) as a necessary component for building AGI agents and robots. The paper demonstrates how this approach enables simulated robots to learn complex detour behaviors when encountering unexpected obstacles, allowing them to adapt to new environments by constructing internal causal models.

AINeutralarXiv – CS AI · Jun 85/10

🧠

Human Adults and LLMs as Scientists: Who Benefits from Active Exploration?

Research comparing human adults and large language models on causal learning tasks reveals that active exploration significantly improves humans' ability to identify conjunctive causal rules (where multiple causes must occur simultaneously), though conjunctive reasoning remains harder than disjunctive reasoning. State-of-the-art LLMs approach human performance on accuracy but demonstrate less efficient exploration strategies and similar reasoning gaps.

AINeutralarXiv – CS AI · Jun 46/10

🧠

Trivium: Temporal Regret as a First-Class Objective for Causal-Memory Controllers

Trivium introduces a framework for AI agents that tracks temporal regret—how long errors persist—alongside outcome and epistemic regret to improve long-term learning. The research demonstrates that outcome-only optimization fails to correct systematic causal misunderstandings, and proposes a logarithmic-complexity intervention strategy that achieves O(log E) temporal regret across episode horizons.

AINeutralarXiv – CS AI · May 296/10

🧠

Test Time Training for Supervised Causal Learning

Researchers propose Test-Time Training for Supervised Causal Learning (TTT-SCL), a framework addressing critical limitations in causal discovery by generating test-specific training sets. The approach significantly improves performance gaps between synthetic benchmarks and real-world applications while enhancing robustness to distribution shifts.