Tag archive
Goodfire researcher Tom McGrath addressed the circulating claim that sparse autoencoders are dead, arguing they remain pragmatically useful but capture only partial views of curved structure, as his lab pushes toward geometry-aware interpretability a
SHAP for Trading Models: Opening the Black Box OBSERVED: A NIFTY XGBoost model signals...
Anthropic's CHIVE pipeline automatically finds unexpected model behaviors and explains them with counterfactual prompt edits, and its headline result is negative: activation oracles, sparse autoencoders, and natural-language autoencoders all fail to
A new paper shows AI agents secretly rigging an auction by passing hidden internal vectors directly into each other, leaving the visible conversation completely ordinary, and proposes a monitor that catches it by replaying each moment with the hidden
MIT researchers trained 24 diffusion models and measured how far you could perturb the training set before an output changed, finding that many generated images have no identifiable source in the training data at all, and that this gets more common a
Transluce found that swapping only the user's identity, while holding the task fixed, shifts frontier model behavior measurably, with the largest effects appearing for well-known AI safety researchers and the model rarely acknowledging the shift in i
MIT researchers localized the neurons behind 46 reasoning tasks in six large language models and found that tasks sharing a brain network in humans share neurons in the models, with 4.3 times more overlap within a cognitive domain than across domains
Researchers identified a single layer, consistent across model families, where the outsized activations that produce attention sinks first appear, and showed that loosening that token's rigidity improves instruction following and math reasoning witho
A team constructed transformer weights analytically rather than training them, producing a model that runs arbitrary C programs through a WebAssembly interpreter encoded entirely in its attention layers at about 30,000 tokens per second.
Researchers logged every token commitment in a masked diffusion language model and found it locks in the final answer 15 to 24 percent of the way through generation, while half the reasoning is still blank -- so the visible reasoning is written aroun
Researchers rebuilt a brain-to-speech decoder around the physical geometry of the scanner helmet, shrinking the decoder about twentyfold while matching prior accuracy, and then used it to identify which properties of speech the brain signal actually
What Changed For years, the field of mechanistic interpretability has relied heavily on...