About
Research & Development
MoE-MLA-RoPE: A New Blueprint for Efficient Small Language Models
TLDR: MoE-MLA-RoPE is a novel AI architecture that combines Mixture of Experts (MoE), Multi-head Latent Attention (MLA), and Rotary Position Embeddings (RoPE) to create...
Research & Development
New Attack Method Boosts Privacy Inferences on Small Language Models
TLDR: A new research paper introduces 'win-k', an improved membership inference attack (MIA) specifically designed for Small Language Models (SLMs). Unlike previous attacks that...
Research & Development
Multi-TW: A New Benchmark for Multimodal AI in Traditional Chinese
TLDR: Multi-TW is the first benchmark designed to evaluate Multimodal Large Language Models (MLLMs) on Traditional Chinese question answering, focusing on both performance and...
Research & Development
BIODISCO: An AI System for Generating and Refining Biomedical Hypotheses
TLDR: BIODISCO is a novel multi-agent AI framework designed to generate and refine biomedical hypotheses. It integrates reasoning from large language models with a...
Research & Development
NaturalGAIA: A New Standard for Evaluating AI Agents in Digital Environments
TLDR: NaturalGAIA is a new benchmark for evaluating AI agents that interact with Graphical User Interfaces (GUIs). It uses "Causal Pathways" to break down...
Research & Development
Optimizing Electric Vehicle Deliveries: A New Framework for Smarter Routing and Charging
TLDR: CARGO is a new framework that co-optimizes electric vehicle (EV) routing and charging for goods delivery. It addresses challenges like limited battery capacity...