# Polarity Composition Hypothesis: Causal Evidence from Negation in NLI

> Source: <https://www.machinebrief.com/news/polarity-composition-hypothesis-causal-evidence-from-negatio-1ixs>
> Published: 2026-11-19 05:00:00+00:00

**Recommendation:** Use Distributed Alignment Search to locate causally effective subspaces for the behavior you care about, then test whether the found representations transfer across input variants and out-of-distribution subsets before relying on them.
More in the playbook "Evals and testing for LLM apps": <https://wpnews.pro/playbooks/evals-testing>

# Polarity Composition Hypothesis: Causal Evidence from Negation in NLI

By Nhi Hoai Doan, Tatsuya Hiraoka, Munachiso Samuel Nwadike, Kentaro InuiSource: 

[arXiv cs.CL](https://arxiv.org/list/cs.CL/recent)
arXiv:2611.23185v1 Announce Type: new 
Abstract: We investigate how language models process negation in natural language 

[inference](https://www.machinebrief.com/glossary/inference)(NLI). We propose the \textit{Polarity Composition Hypothesis}, a high-level causal model of this computation. Behavioral evaluations show that language-model predictions largely conform to the proposed hypothesis. Using Distributed Alignment Search, we identify causally effective subspaces aligned with the variables in the polarity-composition hypothesis. These representations transfer across negation forms and partially to a subset of MNLI. We release the dataset and code.
Get AI news in your inbox

Daily digest of what matters in AI.
