Search papers, labs, and topics across Lattice.
Affiliation:
1
0
3
Scaling and post-hoc interpretability hit a hard wall in agentic AI because models cannot recover causal invariances that observational interaction data never contained in the first place.
