Intrinsic Interpretability of Large Language Models: Key Designs
Explore intrinsic interpretability in large language models with design principles and architectures for transparent, trustworthy AI systems.
Safe Deep Reinforcement Learning for Efficient Heating Control
Optimize building heating with safe deep reinforcement learning to boost energy efficiency, comfort, and demand-side flexibility in smart buildings.
Why Output Diversity Collapses in Post-Trained Models
Explore why output diversity collapses in post-trained language models and how training data impacts varied AI responses in creative tasks.
LogicLoc: Advanced Neurosymbolic Code Localization Framework
Discover LogicLoc, a neurosymbolic framework combining LLMs and Datalog for precise, keyword-agnostic code localization with superior accuracy and efficien...
ToolSimulator: Scalable AI Agent Testing Framework
ToolSimulator enables scalable, safe testing of AI agents using LLM-powered simulations to ensure reliability and early bug detection.