Model Reasoning Analysis
Yasaman and Sameer discuss the importance of understanding how models reason, not just their outputs. They highlight the risk of models memorizing data rather than truly reasoning, urging for more focus on robustness in model evaluation.In this clip
From this podcast
Machine Learning Street Talk (MLST)
#73 - YASAMAN RAZEGHI & Prof. SAMEER SINGH - NLP benchmarks
Related Questions