8F-ai/interactionindex

--- license: mit tags: - human-feedback - preference-modeling - synthetic - coding - safety size_categories: - 1K<n<10K --- # Coding-Safety Preference Index ## Overview This repository contains a synthetic preference dataset built around coding tasks, safety-sensitive refusals, honesty checks, and everyday assistant behavior. It is designed for preference modeling, dataset tooling, and RLHF-style experimentation. ## Layout The repository is organized into four top-level subset folders: - `coding-base` - `coding-online` - `coding-rejection-sampled` - `safety-base` Each folder contains a real gzip-compressed `train.jsonl.gz` file. ## Schema Each line in the data contains a single preference pair with two fields: - `chosen` - `rejected` Both fields use a consistent conversation format: ```json { "chosen": "\n\nHuman: <prompt>\n\nAssistant: <better response>", "rejected": "\n\nHuman: <prompt>\n\nAssistant: <worse response>" } ``` ## Intended Use This dataset is best suited for: - training reward or preference models - testing dataset loaders and conversion pipelines - evaluating instruction-following and refusal behavior - lightweight experimentation with coding and safety-oriented responses ## Notes - The dataset is synthetic and was generated for local experimentation. - The contents emphasize coding help, safety-aware refusal behavior, and honest uncertainty. - Responses are stored in a format compatible with common preference-modeling workflows. ## Loading Example ```python from datasets import load_dataset dataset = load_dataset("json", data_files="coding-base/train.jsonl.gz", split="train") ``` ## Validation The data were checked to ensure: - valid JSONL structure - consistent `chosen` / `rejected` fields - Anthropic-style turn formatting - working gzip compression for subset files许可证：MIT协议 标签： - 人类反馈 - 偏好建模 - 合成数据 - 编码 - 安全 规模类别：1000 < 样本量 < 10000 # 编码安全偏好索引（Coding-Safety Preference Index） ## 概述 本仓库包含一个围绕编码任务、安全敏感拒答、诚实性校验与日常助手交互行为构建的合成偏好数据集，专为偏好建模、数据集工具开发以及人类反馈强化学习（RLHF, Reinforcement Learning from Human Feedback）风格的实验而设计。 ## 数据集结构 本仓库分为四个顶级子集文件夹： - `coding-base` - `coding-online` - `coding-rejection-sampled` - `safety-base` 每个文件夹均包含一个经gzip压缩的`train.jsonl.gz`数据文件。 ## 数据格式规范 数据集中的每一行均为一组偏好对，包含两个字段：`chosen`与`rejected`。两个字段均采用统一的对话格式： json { "chosen": " 人类：<提示词> 助手：<更优回复>", "rejected": " 人类：<提示词> 助手：<较劣回复>" } ## 适用场景 本数据集的适配场景包括： - 训练奖励模型或偏好模型 - 测试数据集加载器与格式转换流水线 - 评估指令遵循与拒答行为 - 面向编码与安全导向回复的轻量型实验 ## 说明 - 本数据集为合成生成，仅用于本地实验场景 - 数据集内容侧重编码协助、安全感知拒答行为以及诚实的不确定性表达 - 回复内容的存储格式兼容主流偏好建模工作流 ## 加载示例 python from datasets import load_dataset dataset = load_dataset("json", data_files="coding-base/train.jsonl.gz", split="train") ## 数据校验 已对数据集进行合规性校验，确保以下要求均满足： - 符合JSONL（JSON Lines）有效结构规范 - 字段`chosen`与`rejected`格式统一 - 采用Anthropic风格的对话回合格式 - 子集文件的gzip压缩格式可正常解压