About
Research & Development
Assessing Multimodal AI’s Counting Abilities in Real-World Scenarios
TLDR: A new benchmark, CountQA, reveals that Multimodal Large Language Models (MLLMs) struggle significantly with accurate object counting in complex, real-world images, achieving a...
Research & Development
Streamlining Knowledge Graph Completion with Probabilistic Circuits
TLDR: A new research paper introduces a framework that uses probabilistic circuits to significantly reduce the number of rules needed for knowledge graph completion,...
Research & Development
GLIDR: Learning Expressive Logic Rules for Knowledge Graphs
TLDR: GLIDR is a new differentiable inductive logic programming (ILP) method that learns more expressive 'graph-like' logical rules, moving beyond the limitations of previous...
Research & Development
PARBALANS: Advancing Mixed-Integer Programming with Parallel Adaptive Search
TLDR: PARBALANS is a novel parallel extension of the BALANS meta-solver designed to accelerate the solution of complex Mixed-Integer Programming (MIP) problems. It achieves...
Research & Development
Building Adaptive and Covert UAV Networks with AI
TLDR: This paper introduces a new framework for Uncrewed Aerial Vehicle (UAV) networks that ensures reliable and covert communication. It combines a generative AI...
Research & Development
Evaluating Multimodal AI on K-12 School Exams
TLDR: MDK12-Bench is a new large-scale benchmark for evaluating Multimodal Large Language Models (MLLMs) using real K-12 school exams across six subjects. It features...