About
Research & Development
New State Space Model Boosts Performance and Speed in Directed Graph Learning
TLDR: DirGraphSSM is a novel graph neural network that extends State Space Models (SSMs) to directed graphs, addressing challenges in capturing long-range causal dependencies...
News & Current Events
Chulalongkorn University Forges Ahead with AI Integration in Education and Research
TLDR: Chulalongkorn University is actively integrating Artificial Intelligence to revolutionize learning and research, as demonstrated by recent key events. A public lecture featuring Amazon...
Research & Development
Advancing Arabic Dialect Identification: A Look at Efficient AI Strategies
TLDR: This research paper explores data-efficient and parameter-efficient strategies for Arabic Dialect Identification (ADI). It finds that Large Language Models (LLMs) struggle to differentiate...
Research & Development
Adaptive Learning for LLMs: A New Approach to Instruction Tuning
TLDR: CAMPUS is a novel instruction tuning framework for Large Language Models (LLMs) that introduces competence-aware, multi-perspective curriculum learning. It dynamically adjusts the training...
Research & Development
Navigating the Complexities of Human-AI Value Alignment: A Comprehensive Review
TLDR: A systematic literature review of 172 articles defines human-AI value alignment as a complex, iterative, and interdisciplinary process. It highlights challenges in expressing...
Research & Development
Optimizing Masked Diffusion Models Through Energy Minimization
TLDR: This research paper introduces a theoretical framework that interprets Masked Diffusion Models (MDMs) as solutions to energy minimization problems in discrete optimal transport....