Causal evidence from representation analysis and interventions shows LLMs use both genuine structure inference and induction circuits in parallel for in-context graph learning.
neurips.cc/paper_files/paper/2022/ha sh/6f1d43d5a82a37e89b0665b33bf3a182-A bstract-Conference.html
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Belief or Circuitry? Causal Evidence for In-Context Graph Learning
Causal evidence from representation analysis and interventions shows LLMs use both genuine structure inference and induction circuits in parallel for in-context graph learning.