Proposes representing LLMs as nondeterministic causal models to enable simpler, implementation-agnostic counterfactual generation compared to prior deterministic approaches.
, t∗ 1), and thus the result follows
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Large Language Models as Nondeterministic Causal Models
Proposes representing LLMs as nondeterministic causal models to enable simpler, implementation-agnostic counterfactual generation compared to prior deterministic approaches.