About
Research & Development
Enhancing Face Anti-Spoofing Generalization Through Multi-View Slot Attention
TLDR: MVP-FAS is a novel face anti-spoofing framework that significantly improves generalization against unseen attacks. It achieves this by introducing Multi-View Slot attention (MVS)...
Research & Development
Boosting MoE-LLM Performance and Speed with Smarter Expert Routing
TLDR: A new post-training strategy called Ban&Pick significantly improves the accuracy and inference speed of Mixture-of-Experts (MoE) Large Language Models (LLMs) without requiring retraining...
Research & Development
Optimizing LLM Performance: Exploring Layer-Wise Scaling for Pre-Training
TLDR: This research introduces three new Layer-Wise Scaling (LWS) variants (Framed, Reverse, Crown) for pre-training Large Language Models (LLMs), which redistribute computational capacity across...
Research & Development
AI-Powered Assistant Streamlines Patient Summaries for Cancer Treatment Decisions
TLDR: The research introduces the Healthcare Agent Orchestrator (HAO), an AI agent system that uses Large Language Models to create accurate and comprehensive patient...
Research & Development
Unlocking Deeper Math Skills in LLMs Through Automated Theorem Proving
TLDR: A new framework generates high-quality, logically sound mathematical reasoning datasets for LLMs by leveraging automated theorem provers and the TPTP axiom library. It...
Research & Development
AI’s Role in Fact-Checking: UNH Explores Fine-Tuning and Prompting for Social Media Claim Extraction
TLDR: The UNH team participated in CheckThat! Task 2, researching methods for extracting verifiable claims from social media posts. They compared fine-tuning LLMs, like...