Search papers, labs, and topics across Lattice.
AIRI, MIPT, ISP RAS Research Center for Trusted AI
1
0
3
LLMs often fail to update their final predictions after interventions on intermediate reasoning steps, suggesting that these structures function more as influential context than stable causal mediators.
