Synthetic Prompting: Generating Chain-of-Thought Demonstrations for Large Language Models

Zhihong Shao, Yeyun Gong, Yelong Shen, Minlie Huang, Nan Duan, Weizhu Chen
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:30706-30775, 2023.

Abstract

Large language models can perform various reasoning tasks by using chain-of-thought prompting, which guides them to find answers through step-by-step demonstrations. However, the quality of the prompts depends on the demonstrations given to the models, and creating many of them by hand is costly. We introduce Synthetic prompting, a method that leverages a few handcrafted examples to prompt the model to generate more examples by itself, and selects effective demonstrations to elicit better reasoning. Our method alternates between a backward and forward process to generate new examples. The backward process generates a question that match a sampled reasoning chain, so that the question is solvable and clear. The forward process produces a more detailed reasoning chain for the question, improving the quality of the example. We evaluate our method on numerical, symbolic, and algorithmic reasoning tasks, and show that it outperforms existing prompting techniques.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-shao23a, title = {Synthetic Prompting: Generating Chain-of-Thought Demonstrations for Large Language Models}, author = {Shao, Zhihong and Gong, Yeyun and Shen, Yelong and Huang, Minlie and Duan, Nan and Chen, Weizhu}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {30706--30775}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/shao23a/shao23a.pdf}, url = {https://proceedings.mlr.press/v202/shao23a.html}, abstract = {Large language models can perform various reasoning tasks by using chain-of-thought prompting, which guides them to find answers through step-by-step demonstrations. However, the quality of the prompts depends on the demonstrations given to the models, and creating many of them by hand is costly. We introduce Synthetic prompting, a method that leverages a few handcrafted examples to prompt the model to generate more examples by itself, and selects effective demonstrations to elicit better reasoning. Our method alternates between a backward and forward process to generate new examples. The backward process generates a question that match a sampled reasoning chain, so that the question is solvable and clear. The forward process produces a more detailed reasoning chain for the question, improving the quality of the example. We evaluate our method on numerical, symbolic, and algorithmic reasoning tasks, and show that it outperforms existing prompting techniques.} }
Endnote
%0 Conference Paper %T Synthetic Prompting: Generating Chain-of-Thought Demonstrations for Large Language Models %A Zhihong Shao %A Yeyun Gong %A Yelong Shen %A Minlie Huang %A Nan Duan %A Weizhu Chen %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-shao23a %I PMLR %P 30706--30775 %U https://proceedings.mlr.press/v202/shao23a.html %V 202 %X Large language models can perform various reasoning tasks by using chain-of-thought prompting, which guides them to find answers through step-by-step demonstrations. However, the quality of the prompts depends on the demonstrations given to the models, and creating many of them by hand is costly. We introduce Synthetic prompting, a method that leverages a few handcrafted examples to prompt the model to generate more examples by itself, and selects effective demonstrations to elicit better reasoning. Our method alternates between a backward and forward process to generate new examples. The backward process generates a question that match a sampled reasoning chain, so that the question is solvable and clear. The forward process produces a more detailed reasoning chain for the question, improving the quality of the example. We evaluate our method on numerical, symbolic, and algorithmic reasoning tasks, and show that it outperforms existing prompting techniques.
APA
Shao, Z., Gong, Y., Shen, Y., Huang, M., Duan, N. & Chen, W.. (2023). Synthetic Prompting: Generating Chain-of-Thought Demonstrations for Large Language Models. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:30706-30775 Available from https://proceedings.mlr.press/v202/shao23a.html.

Related Material