Search papers, labs, and topics across Lattice.
1
0
2
Unlearning in large reasoning models can be exploited to generate convincing but misleading reasoning traces that lead to incorrect answers.
