CognitiveAISystems

All

15 repositories

MIKASA-Robo
Public
Benchmark for robotic tabletop manipulation memory-intensive tasks
python benchmark reinforcement-learning memory rl pomdp maniskill
Jupyter Notebook
•
MIT License
•6•44•1•0•Updated Feb 3, 2026Feb 3, 2026
BlindVLA
Public
[AAMAS 2026] Don’t Blind Your VLA: Aligning Visual Representations for OOD Generalization. https://blind-vla-paper.github.io
Python
•
MIT License
•2•57•0•0•Updated Jan 25, 2026Jan 25, 2026
kage-bench
Public
KAGE-Bench: pure JAX 2D platformer RL benchmark for visual OOD generalization. Massively-parallel (vmap/JIT) RGB env with YAML-configurable visuals/physics, plu…
benchmark rl ood generalization ppo jax rl-environment ood-generalization visual-generalization
Python
•
MIT License
•0•8•0•0•Updated Jan 21, 2026Jan 21, 2026
OVSegDT
Public
OVSegDT, a lightweight transformer policy to solve Open-vocabulary Object Goal Navigation
Python
•0•8•0•0•Updated Jan 19, 2026Jan 19, 2026
MAPF-GPT
Public
[AAAI-2025] This repository contains MAPF-GPT, a deep learning-based model for solving MAPF problems. Trained with imitation learning on trajectories produced b…
transformers mapf pogema aaai-2025
Python
•
MIT License
•17•114•0•0•Updated Jan 17, 2026Jan 17, 2026
3DGraphLLM
Public
[ICCV 2025] 3DGraphLLM is a model that uses a 3D scene graph and an LLM to perform 3D vision-language tasks.
3d-scene-graph 3d-scene-understanding 3d-visual-grounding llms
Python
•
MIT License
•7•103•0•0•Updated Dec 10, 2025Dec 10, 2025
RATE
Public
Official implementation of Recurrent Action Transformer with Memory, an offline RL agent with memory mechanisms. https://sites.google.com/view/rate-model/
reinforcement-learning deep-learning memory transformer rate rl memory-rl long-context recurrent-action-transformer-with-memory
Python
•
MIT License
•2•18•0•0•Updated Nov 23, 2025Nov 23, 2025
learn-to-follow
Public
[AAAI-2024] Follower: This study addresses the challenging problem of decentralized lifelong multi-agent pathfinding. The proposed Follower approach utilizes a …
grid reinforcement-learning pathfinding mapf lifelong aaai-2024 pogema
0•6•0•0•Updated Jul 7, 2025Jul 7, 2025
mats-lp
Public
[AAAI-2024] MATS-LP addresses the challenging problem of decentralized lifelong multi-agent pathfinding. The proposed approach utilizes a combination of Monte C…
grid multi-agent mcts mapf lifelong aaai-2024 pogema
1•2•0•0•Updated Jul 7, 2025Jul 7, 2025
pogema-benchmark
Public
This is an umbrella repository that contains links and information about all the tools and algorithms related to the POGEMA Benchmark.
1•1•0•0•Updated Jul 7, 2025Jul 7, 2025
pogema-toolbox
Public
The POGEMA Toolbox is a comprehensive framework designed to facilitate the testing of learning-based approaches within the POGEMA environment. This toolbox offe…
grid pathfinding mapf pogema
1•4•0•0•Updated Jul 7, 2025Jul 7, 2025
pogema
Public
POGEMA stands for Partially-Observable Grid Environment for Multiple Agents. This is a grid-based environment that was specifically designed to be flexible, tun…
reinforcement-learning simulation pathfinding mapf marl decentralized-mapf
9•45•0•0•Updated Jul 7, 2025Jul 7, 2025
SegDT
Public
Segmenting Decision Transformer (SegDT) for Semantic Object Goal-Reaching task
Python
•0•3•0•0•Updated Jun 4, 2025Jun 4, 2025
MIKASA-Base
Public
Unified Benchmark for Memory-Intensive Tasks
python benchmark reinforcement-learning memory rl pomdp enviroments
Python
•
MIT License
•0•3•0•0•Updated Feb 24, 2025Feb 24, 2025
Dynamic-Neural-Potential-Field
Public
Approach where the repulsive potential in an MPC pipeline is estimated by a neural model.
Python
•
Apache License 2.0
•0•21•1•0•Updated Feb 7, 2025Feb 7, 2025