CoSToM maps ToM features inside LLMs with causal tracing and steers activations in critical layers to boost intrinsic social reasoning and dialogue quality.
InProceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 14397–14413
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
CoSToM:Causal-oriented Steering for Intrinsic Theory-of-Mind Alignment in Large Language Models
CoSToM maps ToM features inside LLMs with causal tracing and steers activations in critical layers to boost intrinsic social reasoning and dialogue quality.