Muhammad Umair Nasir
@utheprodigyn
Followers
279
Following
2K
Media
64
Statuses
2K
👩👩👦 ❤️ || PhD Student at @NYUGameLab and @raillabwits || LLMs x Open-ended Learning || BJJ ||🧍🏽♂️∈ {🇵🇰,🇿🇦}
Johannesburg, South Africa
Joined September 2015
We are very excited to announce our new work: "Word2World: Generating Stories and Worlds through Large Language Models" [ https://t.co/hwnWkK3S3G]. All the thanks to my supervisors Dr. Steven James and Prof. @togelius. Word2World is an LLM-based text-to-env, game-design system.
2
29
141
I have heard AGI has been achieved internally at DSPy.
2
1
12
Now PuzzleScript games can be used for deep learning research through JAX. Perfect to test the new RL algorithm that you’ve been cooking! :)
We introduce PuzzleJAX, a benchmark for reasoning and learning. 🧩💡🦎 PuzzleJAX compiles hundreds of existing grid-based PuzzleScript games to hardware-accelerated JAX environments, and allows researchers to define new tasks via PuzzleScript's concise rewrite rule-based DSL.
0
1
7
There's much more in the paper, of course - check it out! And maybe benchmark your own models? https://t.co/NuAXBpndKY This is work led by @doveliyuchen with contributions from Cong Lin, @utheprodigyn , @_JialinLiu, @FilipoGiovanni, and myself.
0
2
4
Can large language models play simple arcade games? Kind of. Sometimes. Slowly, and not as well as a simple search algorithm. And only if you format the input right. Of course, we made a benchmark to investigate this in more detail, because that's what we do.
4
7
38
The Reinforcement Learning Conference has a good vibe. It feels like people are here because they care about the science, not because it's another line in the CV. I hope this conference doesn't get too big and "prestigious".
2
2
40
The OG Prompt grandmaster.
18
94
2K
Me, talking about games and learning, tomorrow
This week the @cogsci_soc Minds in the Making workshop brings you Vanessa Bermudez and @togelius in conversation about LEARNING 🧠 x DESIGN 🛠️!! And what makes GAMES 🧩awesome for learning. It'll be a blast! Wednesday July 16th 12pm-1pm PT. Register now: https://t.co/TrIQweE8h0.
0
1
6
RL is the technical term for "beatings will continue until morale improves"
12
60
596
I’m so excited to be attending this year. See you in Sweden! #GameAISchool2025
0
0
2
New paper! We are trying to find out how well LLMs can generate functional and novel games in the PuzzleScript game description language, especially when combined with automated playthrough based on search. This is part of our work to create new types of game design assistants.
We've all seen the barrage of video games generated by LLMs on social media. But can we automate this process, and measure the game-generation capabilities of LLMs in a more systematic way? To this end, we introduce ScriptDoctor, a framework for automatically generating
1
9
45
We've all seen the barrage of video games generated by LLMs on social media. But can we automate this process, and measure the game-generation capabilities of LLMs in a more systematic way? To this end, we introduce ScriptDoctor, a framework for automatically generating
2
12
42
Thrilled to share that our paper was just published in JAIR ( https://t.co/awRLUYc7YH)! We formalise task composition in RL using lattice structures📷, building a general framework for logical composition over arbitrary tasks beyond Boolean logic. (1/8) 🧵👇
New Article: "Composition and Zero-Shot Transfer with Lattice Structures in Reinforcement Learning" by Nangue Tasse, James, and Rosman
1
6
21
Could a major opportunity to improve representation in deep learning be hiding in plain sight? Check out our new position paper: Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis. The idea stems from a little-known
47
159
1K
Introducing Continuous Thought Machines New Blog: https://t.co/kLGlwICBDu Modern AI is powerful, but it’s still distinct from human-like flexible intelligence. We believe neural timing is key. Our Continuous Thought Machine is built from the ground up to use neural dynamics as
37
289
1K
The objective paradox in action.
it's interesting to see the big AI labs (at least OpenAI, anthropic, google, xai?) converge on EXACTLY the same extremely specific list of products: - a multimodal chatbot - with a long-compute 'reasoning' mode - and something like "deep research" reminds me of a few years
1
2
31
Remember our work on Word2World, where we generate playable worlds based on stories? Continuing this line of research, we have been working on generating Minecraft environments that tell playable stories. Check it out 👇
🚀Excited to share our new research: Word2Minecraft ( https://t.co/eD3ppTCNdP), which is an improvement to Word2World ( https://t.co/GRXBE3OD1v). Thanks to my supervisors Prof. @togelius, Dr. Steven James and Muhammad @utheprodigyn for their guidance and support! 1/13
1
9
51
🚀Excited to share our new research: Word2Minecraft ( https://t.co/eD3ppTCNdP), which is an improvement to Word2World ( https://t.co/GRXBE3OD1v). Thanks to my supervisors Prof. @togelius, Dr. Steven James and Muhammad @utheprodigyn for their guidance and support! 1/13
1
6
19