TAG
#puzzle-generation
0 篇评论 · 4 篇随笔
相关随笔
Shyne et al.: How Far Do Puzzle Solver Loops Match Human Felt Difficulty — Fukai Reads
A logic-grid-puzzle difficulty study by Shyne, Facey & Cooper. Using solver loops (the pass count of a human-style solver) as a difficulty proxy, they generate difficulty-varied puzzles with a quality-diversity algorithm and, in a 63-player study, show solver loops correlate significantly with subjective difficulty (c=0.30, p=0.015).
Simon Tatham's Portable Puzzle Collection (2004) — 40 Puzzles, Freshly Generated for 22 Years
In 2004, British engineer Simon Tatham — also known as the author of PuTTY — released a free collection of single-player logic puzzles. From Sudoku and Minesweeper to the loop-drawing Loopy and the bridge-building Bridges, all 40 puzzles are freshly generated by computer on every run. Written purely to solve a portability annoyance, this hobby project became, twenty years later, the foundation of an AI algorithmic-reasoning benchmark.
McConnell & Zhao：用遗传算法实时生成「恰到好处」难度谜题 — Fukai 的读书笔记
McConnell 与 Zhao 关于使用遗传算法进行自适应谜题生成的论文。将类似 Cosmic Express 的路径谜题，根据记录玩家解题方式的玩家模型，以每题约7秒的速度实时生成，并在18人实验中证明「仅依赖时间指标」的版本在体感难度与进度感方面逊色于其他版本。
Feng等人：AI能创作「出人意料」的国际象棋谜题吗——Fukai 读论文
以Google DeepMind为核心的研究团队开展了一项利用AI生成创意国际象棋谜题的研究。他们以Lichess数据训练生成模型，再通过强化学习微调，将「出人意料」谜题的生成率从0.22%提升至2.5%，约提高了十倍。亮点在于他们如何将创造力量化为机器可测量的数值。
