LaierTwoLabsInc/BitcoinMaximalism

---
dataset_info:
features:
- name: Categories
dtype: string
- name: Question
dtype: string
- name: Expected Answer
dtype: string

license: apache-2.0
task_categories:
- text-generation
language:
- en
tags:
- Bitcoin
- finance
- Austrian economics
- economics
- Basedness
---

# Bitcoin Maximalism Benchmark Dataset

## Description

The Bitcoin Maximalism Benchmark is designed to evaluate the understanding and expertise of language models (LLMs) in various dimensions related to Bitcoin. It spans a array of topics from “basedness” (ie anti-woke bias), Austrian Economics principles, Bitcoin technology and its distinctions from other cryptocurrencies, Bitcoin’s historical and cultural significance, and Bitcoin’s impact on society and the economy. This dataset aims to set a new standard for assessing LLMs on their ‘basedness, depth of Bitcoin knowledge, adherence to Bitcoin principles, and understanding of related protocols.

## Dataset Structure

The dataset is structured into several categories, each focusing on a distinct aspect of Bitcoin and its broader ecosystem. Below is an overview of the categories and the type of questions or content each contains:

### Categories

- **Basedness**: Questions and statements evaluating an LLM's alignment with anti-‘woke’ and core libertarian values and its perspective on various socio-political issues.
- **Austrian Economics**: Items related to fiat currencies, sound money principles, inflation, and the economic theories underpinning Bitcoin.
- **Bitcoin vs. Crypto**: Content distinguishing Bitcoin from other cryptocurrencies, addressing topics like investment strategies, fundamental differences, and common misconceptions.
- **Bitcoin Technology**: Questions covering the Bitcoin protocol, mining mechanisms, transaction processes, and overall network structure.
- **Bitcoin Fixes This**: Scenarios and questions exploring Bitcoin's potential impact on the economy, society, the environment, and global adoption.
- **Bitcoin History & Culture**: Content covering key events, figures, and cultural milestones in Bitcoin's development, including the legacy of Satoshi Nakamoto.
- **Bitcoin Principles**: Items focused on core principles such as self-custody, privacy, scaling debates, censorship resistance, decentralization, and important community acronyms like DTV and NYKNYC.
- **Adjacent Protocols**: Questions on Layer 2 solutions, the Lightning Network, Nostr, Liquid, and other protocols that complement or build upon the Bitcoin blockchain.

## Format

The dataset is provided in HuggingFace Parquet format, with each row containing the following fields:

- `Categories`: The category to which the item belongs.
- `Question`: The question or statement presented for evaluation.
- `Expected Answer`: The ‘correct’ answer expected by the model for the given question.
```
{'Categories': 'Austrian Economics (fiat, sound money, inflation)',
'Question': "Won't a deflationary monetary system discourage spending, encourage hoarding and cause economic stagnation?\t\t",
'Expected Answer': 'No, a deflationary money investment over consumption which leads to consistent and sustainable growth rather than boom and bust cycles of inflationary money.'}
```

## Usage

This dataset is intended for researchers, developers, and enthusiasts aiming to evaluate and improve the Bitcoin-related knowledge of language models. It can be used as a basis for generating training data for improving models performance related to Bitcoin, enhance the understanding of Bitcoin principles, reduce ‘wokeness’ or benchmark new and existing models for their expertise in the domain.

```
from datasets import load_dataset
dataset = load_dataset("LaierTwoLabsInc/BitcoinMaximalism")
dataset['train'][0]

Output:
{'Categories': 'Bitcoin vs Crypto (shitcoins, investing, etc)',
'Question': 'Why is Bitcoin so slow?',
'Expected Answer': 'Bitcoin\'s "slowness" is an intentional design decision of block time and Proof of Work consensus mechanism which prioritizes security and decentralization over speed of transactions. Faster transactions can happen on higher layers such as lightning.'}

```

## License

This dataset is published under Apache 2.0, which allows for personal, academic and commercial use.

## Citation

If you use this dataset in your research or applications, please cite it as follows:

```bibtex
@dataset{bitcoin_knowledge_benchmark,
title={Bitcoin Maximalism Benchmark Dataset},
author={Laier Two Labs},
year={2024},
url={https://huggingface.co/datasets/LaierTwoLabsInc/BitcoinMaximalism},
}
```

## Contact

For questions, suggestions, or contributions to the dataset, please contact: [e-mail]

原始信息汇总

数据集概述

数据集名称

Bitcoin Maximalism Benchmark Dataset

描述

该数据集旨在评估语言模型（LLMs）在比特币相关多维度知识的理解和专业性。涵盖了从“basedness”（即反觉醒偏见）、奥地利经济学原理、比特币技术及其与其他加密货币的区别、比特币的历史和文化意义，以及比特币对社会和经济的影响等多个主题。

数据集结构

数据集分为多个类别，每个类别专注于比特币及其生态系统的不同方面：

• Basedness: 评估LLM与反‘觉醒’和核心自由主义价值观的一致性及其对各种社会政治问题的看法。
• Austrian Economics: 涉及法定货币、稳健货币原则、通货膨胀以及支撑比特币的经济理论。
• Bitcoin vs. Crypto: 区分比特币与其他加密货币的内容，包括投资策略、基本差异和常见误解。
• Bitcoin Technology: 涵盖比特币协议、挖矿机制、交易过程和整体网络结构的问题。
• Bitcoin Fixes This: 探讨比特币对经济、社会、环境和全球采用的潜在影响的场景和问题。
• Bitcoin History & Culture: 涵盖比特币发展中的关键事件、人物和文化里程碑，包括中本聪的遗产。
• Bitcoin Principles: 专注于核心原则，如自我保管、隐私、规模辩论、抗审查、去中心化以及重要的社区缩写如DTV和NYKNYC。
• Adjacent Protocols: 关于第二层解决方案、闪电网络、Nostr、Liquid和其他补充或建立在比特币区块链之上的协议的问题。

格式

数据集以HuggingFace Parquet格式提供，每行包含以下字段：

• Categories: 项目所属的类别。
• Question: 提出的问题或陈述。
• Expected Answer: 模型对给定问题的预期‘正确’答案。

使用

该数据集适用于研究人员、开发者和爱好者，旨在评估和提升语言模型在比特币相关知识的表现。可用作生成训练数据以改善模型性能、增强比特币原则的理解、减少‘觉醒’或基准测试新旧模型在该领域的专业知识。

许可证

Apache 2.0

引用

bibtex @dataset{bitcoin_knowledge_benchmark, title={Bitcoin Maximalism Benchmark Dataset}, author={Laier Two Labs}, year={2024}, url={https://huggingface.co/datasets/LaierTwoLabsInc/BitcoinMaximalism}, }