Prashant Govindarajan

52 posts

Prashant Govindarajan banner
Prashant Govindarajan

Prashant Govindarajan

@prashantg_17

PhD Candidate at @mila_quebec and @ChandarLab | IIT Madras ‘22 ; Interested in AI for Science and Reinforcement Learning

Montréal, Québec Katılım Ekim 2019
282 Takip Edilen249 Takipçiler
Sabitlenmiş Tweet
Prashant Govindarajan
Prashant Govindarajan@prashantg_17·
If you're looking for natural-sounding text annotations of 3D CAD objects to train language models, check out CADmium, our recent paper accepted in TMLR 📢! Enjoyed working on this with @DavideBald42296 and @qfournier2. I'd be glad to hear any feedback or suggestions!
Chandar Lab@ChandarLab

Can LLMs become CAD designers? Check out “CADmium: Fine-Tuning Code Language Models for Text-Driven Sequential CAD Design”, which is now published in Transactions on Machine Learning Research (TMLR), and led by @prashantg_17, @DavideBald42296, and @qfournier2!

English
2
0
5
326
Prashant Govindarajan retweetledi
Mehran Shakerinava
Mehran Shakerinava@MShakerinava·
Want to know the expressivity of Mamba 3? Come by our ICLR poster! Sat, Apr 25 • 3:15 PM – 5:45 PM Pavilion 4 P4-#4409 The Expressive Limits of Diagonal SSMs for State-Tracking Joint work with Behnoush Khavari, Siamak Ravanbakhsh, and @apsarathchandar.
Mehran Shakerinava tweet media
English
1
10
21
1.4K
Prashant Govindarajan retweetledi
Nour Shaheen
Nour Shaheen@nour_shaheen_·
I'm in Rio de Janeiro for ICLR 2026! 🇧🇷 I will be presenting our paper "Is Depth Heterogeneity a Barrier to Model Merging?" (spoiler: not really) at the Workshop for Test-time Updates on Monday April 27th! Please dm if you're there and you'd like to geek out about... (1/2)
Nour Shaheen tweet mediaNour Shaheen tweet media
English
1
6
29
2.5K
Prashant Govindarajan retweetledi
Chandar Lab
Chandar Lab@ChandarLab·
Sugar-Shack with our lab!! 🍁🪵 Saying bye bye to a long, harsh winter and hello to sunny days!!☀️🤩
Chandar Lab tweet mediaChandar Lab tweet media
English
0
4
9
1.1K
Prashant Govindarajan retweetledi
Chandar Lab
Chandar Lab@ChandarLab·
Quoting this from Aidan, who presented this as a talk to the lab last Friday. If you think LLMs are unreliable for coding, this is the missing piece: reliability comes from how you structure the work around the model. The article lays this out very concretely. 🧵↓
Aidan Li@aidanmrli

I wrote an article on agentic coding for beginners after my talk at @apsarathchandar @ChandarLab group. We cover history of AI coding tools, the importance of model harnesses, and general principles in simple research workflows. Feedback is very welcome! aidanli.dev/writing/articl…

English
1
4
3
344
Prashant Govindarajan retweetledi
Darshan Patil
Darshan Patil@dapatil211·
🧬 New paper Scientific datasets evolve as science evolves. With proteins, new sequences get added, annotations get corrected, and noisy entries get curated out. Introducing CoPeP, a continual-pretraining benchmark for protein LMs. Details 🧵 1/n
Darshan Patil tweet media
English
2
29
84
8.5K
Prashant Govindarajan retweetledi
Chandar Lab
Chandar Lab@ChandarLab·
Today, we’re celebrating Gabriele Prato as they reached a major milestone in their PhD journey and defended their thesis. 🎉 It tackles an underexplored limitation of large language models: knowledge consolidation. The thread below explains what this means and why it matters👇
Chandar Lab tweet media
English
1
5
17
2.9K
Prashant Govindarajan retweetledi
Chandar Lab
Chandar Lab@ChandarLab·
Streaming Reinforcement Learning (RL) is a huge challenge: transitions are used once and discarded immediately. This makes agents extremely sample-inefficient. But what if we could "squeeze" more information out of every single frame? Check out our latest paper!
Chandar Lab tweet media
English
1
12
18
2.9K
Prashant Govindarajan retweetledi
Chandar Lab
Chandar Lab@ChandarLab·
‘The Markovian Thinker’, developed by our lab, has been accepted at @iclr_conf! 

This work achieved long reasoning without the quadratic attention tax LLMs reason in chunks with a bounded state, achieving linear compute, constant memory and scaling beyond its training limit!
GIF
English
1
18
66
8.8K
Prashant Govindarajan retweetledi
Chandar Lab
Chandar Lab@ChandarLab·
Excited to share that we have 3 papers accepted at #ICLR2026! 🇧🇷 Our work this year focuses on efficiency and expressivity: deriving theoretical limits for SSMs, achieving linear scaling for reasoning, and modernizing encoder architectures. A summary of our work 👇 🧵
English
1
11
11
3.4K
Prashant Govindarajan
Prashant Govindarajan@prashantg_17·
If you're looking for natural-sounding text annotations of 3D CAD objects to train language models, check out CADmium, our recent paper accepted in TMLR 📢! Enjoyed working on this with @DavideBald42296 and @qfournier2. I'd be glad to hear any feedback or suggestions!
Chandar Lab@ChandarLab

Can LLMs become CAD designers? Check out “CADmium: Fine-Tuning Code Language Models for Text-Driven Sequential CAD Design”, which is now published in Transactions on Machine Learning Research (TMLR), and led by @prashantg_17, @DavideBald42296, and @qfournier2!

English
2
0
5
326
Prashant Govindarajan
Prashant Govindarajan@prashantg_17·
If you're applying for PhD/MSc positions for Fall 2026, I'd highly encourage you to consider applying to @ChandarLab through the Mila supervision process. Please reach out if you have any questions!
Sarath Chandar@apsarathchandar

I am recruiting several graduate students (both MSc and PhD level) for Fall 2026 @ChandarLab! The application deadline is December 01. Please apply through the @Mila_Quebec supervision request process here: mila.quebec/en/prospective…. More details about the recruitment process here: chandar-lab.github.io/join/

English
0
0
4
398
Prashant Govindarajan
Prashant Govindarajan@prashantg_17·
I’ve been working with Mathieu for more than 1.5 years and I can vouch that he’s an excellent researcher and a great mentor. I’m also attending @COLM_conf this time, and would love to chat with anyone about LLM post-training, reinforcement learning, and AI for Science!
Sarath Chandar@apsarathchandar

If you are attending @COLM_conf and looking to hire a research scientist, I highly recommend you talk to my postdoc, Mathieu Reymond, who is in the job market and at the conference! Mathieu is an expert in mult-objective RL, multi-agent RL, RL for scientific discovery, and RL for LLMs. He is an exceptional researcher with very strong programming/engineering skills (great combo!). Website: mathieu-reymond.github.io You can reach him through the Whova app!

English
0
2
4
1.5K
Prashant Govindarajan retweetledi
Sarath Chandar
Sarath Chandar@apsarathchandar·
If you are attending @COLM_conf and looking to hire a research scientist, I highly recommend you talk to my postdoc, Mathieu Reymond, who is in the job market and at the conference! Mathieu is an expert in mult-objective RL, multi-agent RL, RL for scientific discovery, and RL for LLMs. He is an exceptional researcher with very strong programming/engineering skills (great combo!). Website: mathieu-reymond.github.io You can reach him through the Whova app!
Sarath Chandar tweet media
English
2
12
37
6.5K
Prashant Govindarajan retweetledi
Artem Zholus
Artem Zholus@artemZholus·
Attending #COLM2025 today in Montreal! Shoot me a message if you want to chat about Multimodal LLMs and World Models!
English
0
2
6
3.4K
Prashant Govindarajan
Prashant Govindarajan@prashantg_17·
If you or someone you know from an underrepresented community is planning to apply to graduate school and would benefit from feedback on their application, please check out this application assistance program from our lab.
Sarath Chandar@apsarathchandar

At @ChandarLab, we are happy to announce the third edition of our assistance program to provide feedback for members of communities underrepresented in AI who want to apply to high-profile graduate programs. Want feedback? Details: chandar-lab.github.io/grad_app/. Deadline: Nov 01! cc: @Mila_Quebec, @polymtl, @CIFAR_News

English
0
0
0
183

Keşfet

@apsarathchandar @ChandarLab @Mila_Quebec @polymtl @iclr_conf @DavideBald42296 @qfournier2 @COLM_conf