Search papers, labs, and topics across Lattice.
1
0
2
Selected features from sparse autoencoders can causally steer language models toward desired behaviors, like refusal, revealing new avenues for interpretability and control.
