Calc-X and Calcformers: Empowering Arithmetical Chain-of-Thought through Interaction with Symbolic Systems

Warning

This publication doesn't include Faculty of Education. It includes Faculty of Informatics. Official publication website can be found on muni.cz.
Authors

KADLČÍK Marek ŠTEFÁNIK Michal SOTOLÁŘ Ondřej MARTINEK Vlastimil

Year of publication 2023
Type Article in Proceedings
Conference Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: Main track
MU Faculty or unit

Faculty of Informatics

Citation
Web Manuscript in proceedings
Doi http://dx.doi.org/10.18653/v1/2023.emnlp-main.742
Keywords language models; dataset; arithmetic reasoning; multistep reasoning
Description Despite outstanding performance on many generation tasks, language models are notoriously inclined to make factual errors in tasks requiring arithmetic reasoning. To enable language models to circumvent this deficiency and offload critical computation to a symbolic system, we create a collection of Calc-X datasets that demonstrates the appropriate use of a calculator in reasoning chains. We survey and unify several existing chain-of-thoughts datasets into a proposed novel format, resulting in a standard collection of over 300,000 samples requiring arithmetic reasoning. Finally, we use the new collection to train open-source calculator-assisted language models and show that models trained on Calc-X almost double the accuracy of generating correct results compared to baselines. We make all Calc-X datasets and models publicly available.
Related projects:

You are running an old browser version. We recommend updating your browser to its latest version.