按年被提到
20222023 · 2 次2026
大模型与语言模型›大模型推理与思维链·arXiv:2211.12588 · cs.CL

Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks

Wenhu Chen, Xueguang Ma, Xinyi Wang, William W. Cohen · 2022
开始阅读 ↗PDFalphaXiv 讨论
被提及 · 2 人首次 2023-01-11 · 最近 2025-11-30
2不同的人
2活跃月份
2讨论串
356被引(OpenAlex)
arXiv 摘要(候选层不写导读)

Recently, there has been significant progress in teaching language models to perform step-by-step reasoning to solve complex numerical reasoning tasks. Chain-of-thoughts prompting (CoT) is by far the state-of-art method for these tasks. CoT uses language models to perform both reasoning and computation in the multi-step `thought' process. To disentangle computation from reasoning, we propose `Program of Thoughts' (PoT), which uses language models (mainly Codex) to express the reasoning process as a program. The computation is relegated to an external computer, which executes the generated programs to derive the answer. We evaluate PoT on five math word problem datasets (GSM, AQuA, SVAMP, TabMWP, MultiArith) and three financial-QA datasets (FinQA, ConvFinQA, TATQA) for both few-shot and zero-shot setups. Under both few-shot and zero-shot settings, PoT can show an average performance gain over CoT by around 12\% across all the evaluated datasets. By combining PoT with self-consistency decoding, we can achieve SoTA performance on all math problem datasets and near-SoTA performance on financial datasets. All of our data and code are released in Github https://github.com/wenhuchen/Program-of-Thoughts

候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。

HN 书评

推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及
评论zone411 · 2023-01

“We should see a nice jump in mathematical reasoning of the 2023 generation of chat systems when the model is enabled to write and execute external mini programs and use them in its reasoning. E.g. https://arxiv.org/abs/2211.12588”

Wolfram Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT原评论 ↗
评论zone411 · 2023-01

“It's possible to integrate them to enhance mathematical abilities of LLMs e.g. https://arxiv.org/abs/2211.12588. It's also possible to create fact-grounded retrieval-enhanced language models e.g. https://proceedings.mlr.press/v162/borgeaud22a.html.”

Wolfram Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT原评论 ↗
被提交为 story · 1 次