About
Research & Development
Earth AI: A Unified Approach to Planetary Analysis with Advanced AI Models
TLDR: Earth AI is a new framework that uses a family of specialized AI models (Imagery, Population, Environment) and a Gemini-powered reasoning agent to...
Research & Development
Bridging LLM Flexibility and Rule-Based Reliability for RTS Games
TLDR: Memory-Augmented State Machine Prompting (MASMP) is a new framework for LLM agents in real-time strategy games like StarCraft II. It combines natural language-driven...
Research & Development
AI Framework Enhances Urban Waterlogging Monitoring with Automated Assessment and Early Warnings
TLDR: UWAssess is a new AI framework that uses a combination of vision and vision-language foundation models to automatically detect urban waterlogging from surveillance...
Research & Development
Enhancing LLM Social Intelligence Through Probabilistic Intent Modeling
TLDR: A new framework called Stochastic Theory-of-Mind (SToM) improves Large Language Model (LLM) agents' social intelligence in multi-turn dialogues. It works by maintaining and...
Research & Development
Unlocking GUI Agent Potential: New Benchmark Reveals True Capabilities
TLDR: A new research paper introduces AndroidControl-Curated, a refined benchmark for GUI agents, revealing that previous low performance scores were due to benchmark flaws...
Research & Development
Optimizing Information Retrieval for AI: A Bandit Approach to Complex Queries
TLDR: This research introduces a multi-armed bandit framework for Retrieval-Augmented Generation (RAG) systems to efficiently handle complex user queries. By treating sub-queries as "arms"...