About
Research & Development
Improving Safety and Interpretability in Autonomous Driving Through Entropy Regularization
TLDR: This paper introduces E_loss, a novel entropy-based regularizer for deep perception networks in autonomous driving. By treating networks as hierarchical communication chains and...
Research & Development
CoUn: A Smarter Way for AI Models to Forget Specific Data
TLDR: CoUn is a novel machine unlearning framework that enhances the removal of specific 'forget' data from trained models while preserving knowledge of 'retain'...
Research & Development
Improving Building Efficiency with Causal AI: Introducing GRID
TLDR: GRID is a novel AI framework that leverages a three-stage causal discovery pipeline, combining constraint-based search, neural structural equation modeling, and language model...
Research & Development
Pico: A Systematic Framework for Advancing Small Language Model Research
TLDR: Pico is a new modular framework, consisting of pico-train and pico-analyze, designed to enable systematic, hypothesis-driven research for small and medium-scale language models....
Research & Development
Securing Audio with Light: A Lensless Camera Approach to Encryption and Authentication
TLDR: LenslessMic is a novel audio encryption and authentication system that uses lensless computational imaging to provide a physical layer of security. It converts...
Research & Development
AHA: Real-Time Highlight Detection for Streaming Video Without Future Context
TLDR: AHA is a novel autoregressive framework for online highlight detection in continuous video streams. It predicts frame relevance against a natural language task...