Camila Blank

@camila_blank

math @stanford, @matsprogram

following
followers

WorkspaceBench introduced to evaluate AI interpretability tools

Its creators describe WorkspaceBench as a test of how well interpretability tools reveal important intermediate variables during a model’s forward pass—the computation that produces an output.

2 Sources, 27d ago, first seen 27d ago

OLMo-3-7B Sycophancy Rate Rose During DPO Training

Researcher shared observation on X about model behavior shift in training.

4 Sources, 48d ago, first seen 48d ago

Featured in

WorkspaceBench introduced to evaluate AI interpretability tools

OLMo-3-7B Sycophancy Rate Rose During DPO Training

Featured in

WorkspaceBench introduced to evaluate AI interpretability tools

OLMo-3-7B Sycophancy Rate Rose During DPO Training