Shared biases across LLMs from common pretraining misalign with teaching quality and negatively correlate with intended student learning outcomes, with model ensembles amplifying the misalignment.
hub
ISSN: 2692-8205 Pages: 2025.10.16.679418 Section: New Results
1 Pith paper cite this work, alongside 2,479 external citations. Polarity classification is still indexing.
1
Pith paper citing it
2,479
external citations · external index
hub tools
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Knowledge without Wisdom: Measuring Misalignment between LLMs and Intended Impact
Shared biases across LLMs from common pretraining misalign with teaching quality and negatively correlate with intended student learning outcomes, with model ensembles amplifying the misalignment.