A*STAR
Singapore Agency for Science, Technology and Research; primary engine for foundational and applied AI research
NUS
National University of Singapore; among Asia's top AI research universities. Launched the NUS AI Institute in March 2024, bringing together foundational AI, applied AI and societal impact research
NTU
Nanyang Technological University; major hub for AI and data science research
SMU
Singapore Management University; AI applications in business and society
SUTD
Singapore University of Technology and Design; innovation at the intersection of AI and design
PINNACLE: PINN Adaptive ColLocation and Experimental Points Selection
PINNACLE is an adaptive point selection method that improves Physics Informed Neural Networks (PINNs) training efficiency by automatically optimizing the selection of all training point types using empirical Neural Tangent Kernel theory. The method outperforms existing benchmarks across multiple problem types including forward problems, inverse problems, and transfer learning applications.
BadEdit: Backdooring Large Language Models By Model Editing
AI Singapore researchers present BadEdit, a novel framework for backdoor injection into pre-trained large language models through direct parameter manipulation. The method requires only 15 poisoned samples and 120 seconds to successfully inject backdoors with near 100% attack success rate while minimizing side effects on clean data. BadEdit addresses limitations of traditional weight poisoning methods by significantly reducing data requirements and computational overhead, while demonstrating versatility across diverse task domains including text classification, fact-checking, and conversational sentiment generation.
Utilizing Symbolic Regression to discover a larger class of splits for Decision Trees
This research introduces Symbolic Regression Enhanced Decision Tree (SREDT), which leverages symbolic regression to discover non-linear and multivariate splitting rules for decision trees. Compared to conventional decision trees, SREDT demonstrates superior prediction performance, more compact tree structures, faster inference time, and robustness to noise. The method uses genetic programming to search for closed-form analytical expressions, significantly outperforming standard decision trees on classification tasks.
