Debiasing-DPO reduces bias to spurious social contexts by 84% and improves predictive accuracy by 52% on average for LLMs evaluating U.S. classroom transcripts.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.AI 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
A framework positions classroom interaction studies on axes of scale, duration, and modality to show how design choices affect operationalization, visible mechanisms, and translation to practice, with AI as a way to expand options.
citing papers explorer
-
Mitigating LLM biases toward spurious social contexts using direct preference optimization
Debiasing-DPO reduces bias to spurious social contexts by 84% and improves predictive accuracy by 52% on average for LLMs evaluating U.S. classroom transcripts.
-
Mapping the Methodological Space of Classroom Interaction Research: Scale, Duration, and Modality in an Age of AI
A framework positions classroom interaction studies on axes of scale, duration, and modality to show how design choices affect operationalization, visible mechanisms, and translation to practice, with AI as a way to expand options.