David Khosid PRO
davidkh
·
AI & ML interests
Intelligent robotics, reinforcement learning, physical simulations, generative art
Recent Activity
reacted
to
IliaLarchenko's
post
with 🔥
about 1 month ago
I am presenting Decoder-Only Transformer (DOT) Policy a simple Behavioral Control policy that outperforms SOTA models on two simple benchmark tasks:
✅ PushT (pushing an object to a goal) – 84% success on keypoints, 74% on images (previous best: 75% / 69%)
✅ ALOHA Insert (precise bimanual insertion) – 30% success (previous best: ~21%)
The best part? DOT is much smaller (sometimes 100 times less parameters) than previous SOTA models, trains faster, and avoids complexity:
🚫 No generative models (Diffusion, VAE, GANs)
🚫 No discretization/tokenization of actions
🚫 No reinforcement learning or multi-stage training
✅ Just learns from human demos, plain and simple
This is still early — more complex real-life tasks need testing, and no guarantees it will actually work well there, but I think it's interesting to share. Sometimes, simpler approaches can be just as effective (or even better) than complex ones.
🔗 Open-source code and detailed description: https://github.com/IliaLarchenko/dot_policy
Trained models on Hugging Face:
https://huggingface.co/IliaLarchenko/dot_pusht_keypoints
https://huggingface.co/IliaLarchenko/dot_pusht_images
https://huggingface.co/IliaLarchenko/dot_bimanual_insert
updated
a collection
11 months ago
AI-perception
updated
a collection
11 months ago
AI-perception
Organizations
Collections
1
models
16
davidkh/rl_course_vizdoom_health_gathering_supreme
Reinforcement Learning
•
Updated
davidkh/ppo-CartPole-v1
Reinforcement Learning
•
Updated
davidkh/poca-SoccerTwos
Reinforcement Learning
•
Updated
•
25
davidkh/ppo-unity_pyraminds
Reinforcement Learning
•
Updated
•
6
davidkh/ppo-SnowballTarget
Reinforcement Learning
•
Updated
•
6
davidkh/Reinforce-Pixelcopter-PLE-v0
Reinforcement Learning
•
Updated
davidkh/Reinforce-CartPole-v1
Reinforcement Learning
•
Updated
davidkh/dqn-SpaceInvadersNoFrameskip-v4
Reinforcement Learning
•
Updated
•
4
davidkh/q-taxi-v3
Reinforcement Learning
•
Updated
davidkh/q-FrozenLake-v1-4x4-noSlippery
Reinforcement Learning
•
Updated
datasets
0
None public yet