Ihor Vitenko

ML Research Engineer

Publications

Conference Papers

When Can Pre-Trained Policies Be Reused? Inference-Time Policy Composition from Fixed Policy Libraries

Published in Proceedings of the ACM International Conference on Information and Knowledge Management (CIKM), Short Paper Track, 2026

Inference-time composition of pretrained policies from a fixed policy library, without additional environment interaction.

Recommended citation: Ihor Vitenko, Noha Ibrahim, and Sihem Amer-Yahia. 2026. When Can Pre-Trained Policies Be Reused? Inference-Time Policy Composition from Fixed Policy Libraries. In Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM ’26), November 07–11, 2026, Rome, Italy. ACM, New York, NY, USA, 5 pages. https://doi.org/10.1145/3799682.3839904

Methods for Optimizing Reasoning in Small Vision-Language Models Based on Reinforcement Learning

Published in Theoretical and Applied Problems of Physics, Mathematics and Informatics: Proceedings of the XXIV All-Ukrainian Scientific and Practical Conference of Students, Postgraduates and Young Scientists, 2026

Reinforcement learning methods for optimizing reasoning in small vision-language models.

Recommended citation: Vitenko, I. O., & Kussul, N. M. (2026). "Methods for Optimizing Reasoning in Small Vision-Language Models Based on Reinforcement Learning." In Theoretical and Applied Problems of Physics, Mathematics and Informatics: Proceedings of the XXIV All-Ukrainian Scientific and Practical Conference of Students, Postgraduates and Young Scientists, Kyiv, Ukraine, p. 302. Kyiv: Igor Sikorsky Kyiv Polytechnic Institute, Politechnika Publishing House. ISBN 978-966-990-053-1.
Download Paper

Preprints