Causal interventions expose implicit situation models for commonsense language understanding
Accounts of human language processing have long appealed to implicit `situation models'' that enrich comprehension with relevant but unstated world knowledge. Here, we apply causal intervention techniques to recent transformer models to analyze performance on the Winograd Schema Challenge (WSC), where a single context cue shifts interpretation of an ambiguous pronoun. We identify a relatively small circuit of attention heads that are responsible for propagating information from the context word that guides which of the candidate noun phrases the pronoun ultimately attends to. We then compare how this circuit behaves in a closely matched `syntactic'' control where the situation model is not strictly necessary. These analyses suggest distinct pathways through which implicit situation models are constructed to guide pronoun resolution.
Code (1)
Tasks
World KnowledgeSimilar Papers 제목 키워드 기반
GLUCOSE: GeneraLized and COntextualized Story Explanations
When humans read or listen, they make implicit commonsense inferences that frame their understanding of what happened and why. As a step toward AI systems that can build similar mental models, we introduce GLUCOSE, a lar…
Disentangling Mixtures of Unknown Causal Interventions
In many real-world scenarios, such as gene knockout experiments, targeted interventions are often accompanied by unknown interventions at off-target sites. Moreover, different units can get randomly exposed to different …
Implicit Causal Representation Learning via Switchable Mechanisms
Learning causal representations from observational and interventional data in the absence of known ground-truth graph structures necessitates implicit latent causal representation learning. Implicit learning of causal me…
DisentanglementRepresentation LearningA Critique on the Interventional Detection of Causal Relationships
Interventions are of fundamental importance in Pearl's probabilistic causality regime. In this paper, we will inspect how interventions influence the interpretation of causation in causal models in specific situation. To…
validKnowledge-Augmented Language Models for Cause-Effect Relation Classification
Previous studies have shown the efficacy of knowledge augmentation methods in pretrained language models. However, these methods behave differently across domains and downstream tasks. In this work, we investigate the au…
Cause-Effect Relation ClassificationClassificationCommonsense Causal ReasoningLanguage Modelling+2