构建互动叙事中用户偏好数据集，揭示大模型在个性化选择上的短板。
Rushes: A Human Preference Dataset for Pluralistic Alignment
- 通过游戏界面收集用户在分支剧情中的实时选择行为
- 大模型预测准确率仅34.23%，低于流行度基线的36.4%
- 适合研究个性化生成与多目标对齐的学者和开发者
我们提出Rushes，一个用于研究交互式叙事环境中用户真实参与偏好的数据集与基准。Rushes通过游戏界面收集用户在AI生成的分支叙事中，从有限候选集中做出选择的行为数据，记录完整候选集、用户选择及演进的叙事上下文，形成带时间顺序、持久用户标识的轨迹。该数据集包含来自8,167名用户的44,226个决策事件，覆盖六款游戏，捕捉序列化、个性化的参与行为，而非静态判断。我们发现用户选择呈现结构化、非随机模式，其选择熵远低于均匀分布基准。我们将Rushes定位为多元对齐的诊断基准，揭示显著的参与差距：当前最先进的大模型（包括GPT-5）在事件级选择预测上未能超越简单基线。经典矩阵分解（SVD）可捕获37.7%的个性化信号，而前沿大模型仅达34.23%，甚至无法超越流行度基线（36.4%）。这表明，单一全局目标（如现代RLHF所用）难以捕捉异质、依赖上下文的参与信号，导致高性能模型仍趋向于主流偏好，忽视个体轨迹。我们公开Rushes，以支持生成系统中多元对齐与序列决策的研究。完整平台与数据代码将发布于：https://github.com/microsoft/rushes
原文摘要 · Abstract (English)
We introduce Rushes, a dataset and benchmark for studying revealed human engagement preferences in interactive narrative environments. Rushes is collected through a game interface where users interact with AI-generated branching narratives and select one choice from a small, explicit candidate set at each decision point. Each interaction logs the full candidate set, the user's choice, and the evolving narrative context, yielding time-ordered trajectories with persistent user-level identifiers. Rushes contains 44,226 decision events from 8,167 unique users across six games, capturing sequential, personalized engagement behavior rather than static judgments. We show that user choices exhibit structured, non-random patterns, quantified by a low choice entropy relative to a uniform baseline. We position Rushes as a diagnostic benchmark for pluralistic alignment and demonstrate a robust Engagement Gap: state-of-the-art LLMs, including GPT-5, fail to outperform simple baselines. While classical Matrix Factorization (SVD) captures measurable personalized signal (37.7%), frontier LLMs (34.23%) struggle to even match the Popularity Baseline (36.4%) on event-level choice prediction. This gap suggests that single, population-level objectives, like those used in modern RLHF, appear insufficient to capture heterogeneous, context-dependent engagement signals. As a result, even highly capable models default to majority preferences rather than adapting to individual trajectories. We release Rushes to support research into pluralistic alignment and sequential decision-making in generative systems. The full code for the platform and dataset will be available here: https://github.com/microsoft/rushes
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。