Playing with RL, World Models + GPUs, CS undergrad. https://xitter.cf/t.co/GDkIxkjQAQ, https://xitter.cf/t.co/IFYkVC5d9Y, GSoC mentee @JuliaLanguage

Mumbai
Joined August 2023
Param Thakkar retweeted
Replying to @lucascassiano
It was never about the ability to generate environments. The term got hijacked by Silicon Valley shot chasers. It’s about the ability to imagine complex dynamic future states off current information in a (relatively) fixed cost compared to paying the exponential cost to simulate
1
2
1
46
995
Param Thakkar retweeted
How do I join a lab that does this
We are bringing agents to the wet-lab bench at Google Research to navigate complex biological landscapes. Learn how we are grounding agents in physical lab robotics to run fast closed-loop molecular design cycles, ultimately discovering single-digit nanomolar protein binders for cancer targets →goo.gle/4B5mMMB
24
18
1
681
44,247
Param Thakkar retweeted
The games we've long wanted, which didn't exist because they'd cost several times more to make than they'd earn back, will soon become possible. As will the games developers wanted to build, but couldn't with the resources available. It'll be a new golden age of gaming.
2,157
1,307
916
17,626
5,689,111
Param Thakkar retweeted
twitter is the bloomberg of ai research
136
294
34
4,762
102,801
My first PR to Blender just got merged 🎉 Blender is genuinely one of the best open source projects out there, and it’s a great feeling to have contributed something to it. Thanks to the reviewers for the feedback along the way. More to come. projects.blender.org/blender…
1
7
143
Param Thakkar retweeted
Do a kickflip!
4,510
25,351
10,868
207,095
27,909,449
Param Thakkar retweeted
Real life, baby
19,283
96,581
45,625
776,754
97,315,668
Param Thakkar retweeted
Turns out the future does have flying cars
48,095
100,245
32,674
834,381
78,381,825
Param Thakkar retweeted
deedy is right. india treats its brightest minds like shit and then cries about brain drain. why should talented people stay in a country that consistently punishes competence and rewards mediocrity?
The smartest Indians continue to come to build in the US for the simple reason that the economic incentives are far far better. Perplexity ($20B), Skild ($14B+), Baseten ($13B), Together ($8B), Glean ($8B), Modal ($5B), Rubrik ($25B), Harness ($5B), Blacksmith, Parallel, Greptile, CodeRabbit, Composio, Distyl ($2B), Resolve, Luma, Cartesia, Wispr, dMatrix, Arize and many more - all started by at least one Indian-born founder. The US is a far easier to raise capital in with far less dilution. Most Indian companies are switching where they're domiciled to get access to the US capital markets. Every single Indian venture firm is trying very hard to invest in the US or risk not participating in the AI race (ironically, the management fees of venture firms are so high in India relatively that this is the one group that net of carry actually has incentive to be in India because they take home well north of a living wage whether or not funds perform do) While Zomato ($35B now, $8B at IPO) might be the biggest outcome in Indian venture, Cursor ($60B), Reflection (potentially $20B+), World Labs ($8B), OpenRouter ($8B), HuggingFace ($12B) and of course the $1T+ big labs quickly make those outcomes seem small and happen on a much smaller time scale, with much higher IRR. Engineers at these startups regularly get paid $1M/yr, and make $10M+ on exit. Even in BigTech, it is trivial to pull up levels fyi and see that if you're L6/L7+ with ~10 years of experience, you're making $1M/yr there too. And of course, we all know the slew of Indian CEOs of companies here, not to mention a huge bench of lesser known executives across the board in public and private companies. I say all this because every time there is a slight immigration hiccup in the US, I hear the online pleas to "build in India", I struggle to understand why most engineers, academics or founders would do that net of desire to be close to family, failure to obtain visa and devout patriotism. This is not a superiority thing or a US v India thing. The economic incentives at least for technologists to being in the US outweigh India by at least an order of magnitude.
48
223
13
2,558
40,011
Just received my copy of The Infinity Machine 📖 deepmind lore time 🫡
7
81
Param Thakkar retweeted
waking up and seeing your agent stopped to ask for permission 8h ago
310
2,538
259
45,275
1,019,162
Param Thakkar retweeted
Playing video games is associated with higher intelligence.
215
1,402
328
11,491
991,612
Google Summer of Code 2026, Completed Successfully 😃🎉
3
31
680
Param Thakkar retweeted
We’re teaching a new course at Penn: CIS 6280 · World Models 🌍 cis.upenn.edu/~cis6280/ Every day, LLMs amaze me with one more thing they can solve. But I’ve always felt there’s more to learn than what humans have written down—from observing the world, interacting with it, and experiencing what happens next. To me, that’s what world models are about—and why I see them as one of AI’s next big bets. However, I struggled to find a systematic course on them—so we are building one! Topics will include: representation learning, generative models, simulation, model-based RL, video and 3D generation, world models for robotics, reasoning, and code-based world models. Hands-on work will include: - Building an environment - Training a world model - Learning a policy and a final research project with leaderboard. We’re already six lectures in. Slides, demos, and readings are public, and we’ll keep adding materials throughout the fall. Big thanks to our TA team— @TongMutianTMT @hagsaeng_bag @EnxinSong @KeelyAi04 —for helping bring this course to life!
56
368
26
2,396
576,080
Excited to share: our paper "Modeling What Changes: Sparse, Residual World Models for Object-Centric Manipulation" was accepted to the WORLDS Workshop @ IROS 2026! 🤖🌎 A thread on what we found including the parts that didn't work the way we first hoped 🧵👇
2
6
145
Why write it this way? Because "predicts dynamics far more accurately" is a claim our own numbers don't support, and we say so directly in the paper. What we can defend: a change detector that avoids error injection, not an accurate mover predictor.
1
1
9