Abstract
While modern language models demonstrate impressive generative capabilities, they often struggle with constrained, counter-intuitive creative tasks. To address this limitation, we explore chess puzzle generation as a rigorous testbed for computational creativity and reasoning, a domain where altering a single piece can invalidate an entire solution. We propose a novel approach for conditional generation of creative chess puzzles using masked diffusion models. Unlike previous methods, our non-directional diffusion approach allows for conditioning on specific tactical themes and partial board positions. We introduce a novel auxiliary task of simultaneous best-move prediction, which improves solution uniqueness by 11.6% and theme-conditioning accuracy by 2.5%. To further optimize solution uniqueness and theme conditioning, we establish a reinforcement learning framework adapted from Denoising Diffusion Policy Optimization (DDPO). This RL training increases the yield of unique and theme-matching positions by 89.1%. Finally, we release the first open-weights models (Appendix B) for chess puzzle generation, offering a new pathway for controllable, creative generation.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Selkee, A., Rissanen, S., Feng, X., Zahavy, T., & Malmi, E. (2026). Conditional Generation of Creative Chess Puzzles with Diffusion Models. https://omanscience.com/en/articles/conditional-generation-of-creative-chess-puzzles-with-diffusion-models
MLA 9
Selkee, Aatu, et al. "Conditional Generation of Creative Chess Puzzles with Diffusion Models." https://omanscience.com/en/articles/conditional-generation-of-creative-chess-puzzles-with-diffusion-models.
Chicago (author–date)
Selkee, Aatu, Severi Rissanen, Xidong Feng, Tom Zahavy, and Eric Malmi. 2026. "Conditional Generation of Creative Chess Puzzles with Diffusion Models." https://omanscience.com/en/articles/conditional-generation-of-creative-chess-puzzles-with-diffusion-models.
Harvard
Selkee, A., Rissanen, S., Feng, X., Zahavy, T. and Malmi, E. (2026) 'Conditional Generation of Creative Chess Puzzles with Diffusion Models', Available at: https://omanscience.com/en/articles/conditional-generation-of-creative-chess-puzzles-with-diffusion-models.
Vancouver
Selkee A, Rissanen S, Feng X, Zahavy T, Malmi E. Conditional Generation of Creative Chess Puzzles with Diffusion Models. https://omanscience.com/en/articles/conditional-generation-of-creative-chess-puzzles-with-diffusion-models
IEEE
A. Selkee, S. Rissanen, X. Feng, T. Zahavy, and E. Malmi, "Conditional Generation of Creative Chess Puzzles with Diffusion Models," https://omanscience.com/en/articles/conditional-generation-of-creative-chess-puzzles-with-diffusion-models.