Papers by Vladislav Kruglikov
T-pro 2.0: An Efficient Russian Hybrid-Reasoning Model and Playground (2026.eacl-demo)
Copied to clipboard
Dmitrii Stoianov, Danil Taranets, Olga Tsymboi, Ramil Latypov, Almaz Dautov, Vladislav Kruglikov, Surkov Nikita, German Abramov, Pavel Gein, Dmitry Abulkhanov, Mikhail Gashkov, Viktor Zelenkovskiy, Artem Batalov, Aleksandr Medvedev, Anatolii Potapov
| Challenge: | Recent foundation models show that reasoningoriented training and improved decoding methods can substantially boost both accuracy and speed. |
| Approach: | They propose an open-weight Russian LLM for hybrid reasoning and efficient inference. |
| Outcome: | The proposed model supports direct answering and reasoning-trace generation . the model and inference pipeline can be extended or modified to suit Russian-language reasoning . |