Vindula Jayawardana
@vindulaj.bsky.social
Research Scientist at Toyota Research Institute | MIT PhD
Finally, this work is out!!
We found that controlling vehicle speeds to mitigate traffic across a city can cut carbon emissions between 11 and 22 percent. To do this, we used deep reinforcement learning to optimize one million eco-driving scenarios. 🚗🤖🧠
We found that controlling vehicle speeds to mitigate traffic across a city can cut carbon emissions between 11 and 22 percent. To do this, we used deep reinforcement learning to optimize one million eco-driving scenarios. 🚗🤖🧠
This project was 4 years in the making and it's finally out!
We found that controlling vehicle speeds to mitigate traffic across a city can cut carbon emissions between 11 and 22 percent. To do this, we used deep reinforcement learning to optimize one million eco-driving scenarios. 🚗🤖🧠
We found that controlling vehicle speeds to mitigate traffic across a city can cut carbon emissions between 11 and 22 percent. To do this, we used deep reinforcement learning to optimize one million eco-driving scenarios. 🚗🤖🧠
August 16, 2025 at 6:35 PM
Finally, this work is out!!
We found that controlling vehicle speeds to mitigate traffic across a city can cut carbon emissions between 11 and 22 percent. To do this, we used deep reinforcement learning to optimize one million eco-driving scenarios. 🚗🤖🧠
We found that controlling vehicle speeds to mitigate traffic across a city can cut carbon emissions between 11 and 22 percent. To do this, we used deep reinforcement learning to optimize one million eco-driving scenarios. 🚗🤖🧠
Checkout our recent work on cooperative eco-driving with deep RL at city scale. #reinforcementlearning #eco-driving #autonomousvehicles
📣 I will be giving two talks this week at the TRB DATA conference in Seattle, which explores the intersection of transportation, artificial Intelligence, and data analysis. I love these practitioner + researcher conferences. 🤖🧠👇
May 26, 2025 at 8:12 PM
Checkout our recent work on cooperative eco-driving with deep RL at city scale. #reinforcementlearning #eco-driving #autonomousvehicles
Check out our recent #ICLR work!
Excited to announce IntersectionZoo, a benchmark that uses a real-world traffic problem to test generalization progress in deep reinforcement learning, particularly multi-agent contextual RL. 🤖🧠
MIT News coverage: news.mit.edu/2025/new-too...
Benchmark: intersectionzoo-docs.readthedocs.io
MIT News coverage: news.mit.edu/2025/new-too...
Benchmark: intersectionzoo-docs.readthedocs.io
.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:23znfzkxo3w72edq2u6huiw6/bafkreiggaxsnhuw625wq3cut4vofiuvfounlsgots3g67s5yllx4kfkvpi@jpeg)
May 5, 2025 at 9:32 PM
Check out our recent #ICLR work!
Reposted by Vindula Jayawardana
We built an AI assistant that plays Minecraft with you.
Start building a house—it figures out what you’re doing and jumps in to help.
This assistant *wasn't* trained with RLHF. Instead, it's powered by *assistance games*, a better path forward for building AI assistants. 🧵
Start building a house—it figures out what you’re doing and jumps in to help.
This assistant *wasn't* trained with RLHF. Instead, it's powered by *assistance games*, a better path forward for building AI assistants. 🧵
April 11, 2025 at 10:17 PM
We built an AI assistant that plays Minecraft with you.
Start building a house—it figures out what you’re doing and jumps in to help.
This assistant *wasn't* trained with RLHF. Instead, it's powered by *assistance games*, a better path forward for building AI assistants. 🧵
Start building a house—it figures out what you’re doing and jumps in to help.
This assistant *wasn't* trained with RLHF. Instead, it's powered by *assistance games*, a better path forward for building AI assistants. 🧵
Reposted by Vindula Jayawardana
We've built a simulated driving agent that we trained on 1.6 billion km of driving with no human data.
It is SOTA on every planning benchmark we tried.
In self-play, it goes 20 years between collisions.
It is SOTA on every planning benchmark we tried.
In self-play, it goes 20 years between collisions.
February 6, 2025 at 6:34 PM
We've built a simulated driving agent that we trained on 1.6 billion km of driving with no human data.
It is SOTA on every planning benchmark we tried.
In self-play, it goes 20 years between collisions.
It is SOTA on every planning benchmark we tried.
In self-play, it goes 20 years between collisions.