Publications

Ku Onoda, Paavo Parmas, Hiroki Furuta, Soichiro Nishimori, Yuta Oshima, Shohei Taniguchi, and Yutaka Matsuo.

Multi-Axis Max@K Reinforcement Learning for Representative Diversity in Text-to-Image Generation

IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2027).

[arxiv] [code]

Ku Onoda, Paavo Parmas, Manato Yaguchi, and Yutaka Matsuo.

Does 'Do Differentiable Simulators Give Better Policy Gradients?' Give Better Policy Gradients?

International Conference on Learning Representations (ICLR 2026).

[arxiv]

Yuta Oshima*, Ku Onoda*, Yusuke Iwasawa, Masahiro Suzuki, Yutaka Matsuo, and Hiroki Furuta. (*Equal contribution.)

AutoRef: Harness Optimization for Agentic Multi-Reference Image Generation

arXiv preprint.

[arxiv] [code]

Ku Onoda, Yuta Oshima, Shohei Taniguchi, Soichiro Nishimori, Paavo Parmas, Hiroki Furuta, and Yutaka Matsuo.

Diversity-Aware Max@k Optimization for Improving Best-of-N Performance in Image Generation with Diffusion Models

The 40th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI 2026), 2026.

[paper]

Ku Onoda, Paavo Parmas, and Yutaka Matsuo.

Efficient and Low Bias Policy Gradient Estimation in Contact Rich Differentiable Simulation

The 39th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI 2025), 2025.

[paper]