PInKS: Preconditioned Commonsense Inference with Weak Supervision

Anonymous

Abstract: Reasoning with preconditions such as "glass can be used for drinking water unless the glass is shattered" remains an open problem for language models. The main challenge lies in the scarcity of preconditions data and the model's lack of support for such reasoning. We present PInKS, Preconditioned Commonsense Inference with Weak Supervision, an improved model for reasoning with preconditions through minimum supervision. We show, both empirically and theoretically, that PInKS improves the results across the benchmarks on reasoning with the preconditions of commonsense knowledge~(up to 0.4 macro-f1 scores). We further investigate the robustness of our method through PAC-Bayesian informativeness analysis, recall measures, and ablation study.

Paper Type: long

Consent To Share Data: yes