About
Research & Development
Boosting LLM Performance Through Automated Data Curation
TLDR: Refine-n-Judge is an automated method that uses a single Large Language Model (LLM) to iteratively refine and judge its own outputs, creating high-quality,...
Research & Development
Adaptive AI Agents: Designing Dynamic Behaviors with Polymorphic Combinatorial Frameworks
TLDR: The Polymorphic Combinatorial Framework (PCF) uses LLMs and mathematical principles (like topos theory) to create adaptive AI agents. It defines agent behaviors through...
Research & Development
LLMs Enter the Pokémon Arena: A New Benchmark for AI Strategy
TLDR: This research introduces "LLM Pokémon League," a novel tournament system that evaluates Large Language Models (LLMs) as strategic agents in Pokémon battles. By...
Research & Development
DRKF: Advancing Emotion Recognition Through Decoupled Representations and Knowledge Fusion
TLDR: DRKF (Decoupled Representations with Knowledge Fusion) is a new method for Multimodal Emotion Recognition (MER) that addresses challenges like modality differences and inconsistent...
Research & Development
New Benchmark Assesses AI’s Quantitative Chemistry Skills
TLDR: QCBench is a new benchmark with 350 quantitative chemistry problems across 7 subfields and 3 difficulty levels, designed to evaluate LLMs' step-by-step numerical...
Research & Development
Evaluating AI’s Tool-Using Prowess: Introducing LiveMCPBench
TLDR: LiveMCPBench is a new benchmark for evaluating large language model (LLM) agents in complex, real-world environments with many tools. It includes a large...