Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding
| Date | Stars |
|---|---|
| 2026-07-31 | 421 |
| 2026-08-06 | 421 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Meta-Prompting [](https://arxiv.org/abs/2401.12954) **Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding**  **Enhancing GPT-4 with meta-prompting.** In this study, we introduce and examine the effectiveness of meta- prompting, contrasting it with a range of zero-shot prompting techniques, including standard zero-shot (Std), zero-shot chain-of-thought (0-CoT), generic and dynamic expert (Ex-St and Ex-Dy), and multipersona (MP). Our research demonstrates that meta-prompting, particularly when combined with a Python interpreter, significantly improves overall accuracy and robustness in GPT-4 across a variety of tasks. ## Table of Contents [**Abstract**](#abstract) | [**Dataset**](#tasks-and-datasets) | [**Prompts**](#prompt-templates) | [**Outputs**](#model-outputs) | [**Implementation and Evaluation**](#running-experiments-and-evaluation) | [**Citation**](#citation-guidelines) | [**Related Work**](#related-and-concurrent-investigations) | [**Thanks**](#acknowledgements) ## Abstract We introduce **meta-prompting**, an effective scaffolding technique designed to enhance the functionality of language models (LMs). This approach transforms a single LM into a multi-faceted conductor, adept at managing and integrating multiple independent LM queries. By employing high-level instructions, meta-prompting guides the LM to deconstruct complex tasks into smaller, more manageable subtasks. These subtasks are then handled by distinct "expert" instances of the same LM, each operating under specific, tailored instructions. Central to this process is the LM itself, in its role as the conductor, which ensures seamless communication and effective integration of the outputs from these expert models. It additionally employs its inherent critical thinking and robust verification processes to refine and authenticate the end result. This collaborative prompting approach empowers a single LM to simultaneously act as a comprehensive orchestrator and a panel of diverse experts, significantly enhancing its performance across a wide array of tasks. The zero-shot, task-agnostic nature of meta-prompting greatly simplifies user interaction by obviating the need for detailed, task-specific instructions. Furthermore, our research demonstrates the seamless integration of external tools, such as a Python interpreter, into the meta-prompting framework, thereby broadening its applicability and utility. Through rigorous experimentation with GPT-4, we establish the superiority of meta-prompting over conventional scaffolding methods: When averaged across all tasks, including the Game of 24, Checkmate-in-One, and Python Programming Puzzles, meta-prompting, augmented with a Python interpreter functionality, surpasses standard prompting by 17.1\%, expert (dynamic) prompting by 17.3\%, and multipersona prompting by 15.2\%. ## Data, Prompt, and Output Files ### Tasks and Datasets All input-output files related to tasks are available for access within the `/data` directory. - If you wish to access our raw input-output data files, they are also available through Hugging Face Datasets at [https://huggingface.co/datasets/turingmachine/meta-prompting](https://huggingface.co/datasets/turingmachine/meta-prompting) ### Prompt Templates All prompt templates and system instructions used in our experiments are located in the `/prompts` directory. ### Model Outputs All outputs generated during our experiments are stored in the `/outputs` directory. - The majority of our experiments were conducted in August 2023. ## Meta-Prompting Implementation If you are interested in implementing our meta-prompting framework, you can find an example implementation in `utils/meta_scaffolding.py`. This code contains the primary scaffolding structure and necessary functionalities required
Excerpt of 11,741 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:20bbcbd8e4383bb8, llm:Repository title/description: 'Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding' — indicates techniques for prompting language models to improve performance across tasks.
matched fp:20bbcbd8e4383bb8, llm:Repository title/description: 'Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding' — indicates techniques for prompting language models to improve performance across tasks.
matched fp:20bbcbd8e4383bb8, llm:Repository title/description: 'Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding' — indicates techniques for prompting language models to improve performance across tasks.