Petr Kuderov

I'm a Research Scientist in Artificial Intelligence and Cognitive Systems at Moscow Institute of Physics and Technology working on bio-inspired approaches to temporal memory and reinforcement learning. My research focuses on developing novel architectures and algorithms that bridge neuroscience-inspired models with practical AI applications.

I work closely with Dr. Aleksandr Panov and colleagues at MIPT, where we're advancing biologically plausible approaches to sequential learning and decision-making. Our recent work explores Hebbian learning principles for spatial encoding, distributed temporal memory systems, and hierarchical planning algorithms that take inspiration from how biological brains process temporal information.

Much of my current research centers on developing more stable and interpretable ways to learn temporal representations, particularly through projects like the Distributed Hebbian Temporal Memory (DHTM) algorithm and applications in instruction following for embodied AI. I'm especially interested in how principles from neuroscience can inform better artificial learning systems while maintaining computational efficiency and practical applicability.

Publications

Hebbian spatial encoder with adaptive sparse connectivity

Petr Kuderov, E. Dzhivelikian, A. Panov

Cognitive Systems Research 2024

ABS
Instruction Following with Goal-Conditioned Reinforcement Learning in Virtual Environments

Instruction Following with Goal-Conditioned Reinforcement Learning in Virtual Environments

Zoya Volovikova, Alexey Skrynnik, Petr Kuderov, Aleksandr Panov

European Conference on Artificial Intelligence 2024

ABSHTMLPDF

Attractor Properties of Spatiotemporal Memory in Effective Sequence Processing Task

Petr Kuderov, E. Dzhivelikian, A. Panov

Optical Memory and Neural Networks 2023

ABSPDF
Learning Successor Features with Distributed Hebbian Temporal Memory

Learning Successor Features with Distributed Hebbian Temporal Memory

E. Dzhivelikian, Petr Kuderov, A. Panov

ABSHTMLPDF

Hierarchical intrinsically motivated agent planning behavior with dreaming in grid environments

E. Dzhivelikian, Artem Latyshev, Petr Kuderov, A. Panov

Brain Informatics 2022

ABSPDF

Soft Adaptive Segments for Bio-Inspired Temporal Memory

Artem Prokhorenko, E. Dzhivelikian, Petr Kuderov, A. Panov

Hybrid Artificial Intelligence Systems 2024

ABS

Interpreting Decision Process in Offline Reinforcement Learning for Interactive Recommendation Systems

Zoya Volovikova, Petr Kuderov, A. Panov

International Conference on Neural Information Processing 2023

ABS

Stabilize Sequential Data Representation via Attraction Module

Petr Kuderov, E. Dzhivelikian, A. Panov

BI 2023

ABS

Stability and Similarity Detection for the Biologically Inspired Temporal Pooler Algorithms

Ivan Rodkin, Petr Kuderov, A. Panov

BICA*AI 2022

ABS
Planning with Hierarchical Temporal Memory for Deterministic Markov Decision Problem

Planning with Hierarchical Temporal Memory for Deterministic Markov Decision Problem

Petr Kuderov, A. Panov

International Conference on Agents and Artificial Intelligence 2021

ABS

Intrinsic Motivation to Learn Action-State Representation with Hierarchical Temporal Memory

E. Dzhivelikian, Artem Latyshev, Petr Kuderov, A. Panov

BI 2021

ABS