EDU-CIRCUIT-HW reveals large latent recognition failures in MLLMs on real handwritten university STEM solutions, limiting auto-grading reliability, though hybrid human-AI routing of only 3.3% cases improves outcomes.
background-color:#ffd6d6;color:#b00020;font-weight:bold
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
EDU-CIRCUIT-HW: Evaluating Multimodal Large Language Models on Real-World University-Level STEM Student Handwritten Solutions
EDU-CIRCUIT-HW reveals large latent recognition failures in MLLMs on real handwritten university STEM solutions, limiting auto-grading reliability, though hybrid human-AI routing of only 3.3% cases improves outcomes.