Liberal partial-credit prompting reduces question-level grading error for all six tested LLMs, with ChatGPT 5.5 Thinking (LIBERAL) achieving the lowest MAE of 1.87.
Using large language models for automated grading of student writing about science,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CY 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LLMs as Teaching Assistants for Mathematics Exam Grading: Reliability, and Practical Usability
Liberal partial-credit prompting reduces question-level grading error for all six tested LLMs, with ChatGPT 5.5 Thinking (LIBERAL) achieving the lowest MAE of 1.87.