Falguni Das Shuvo

Machine Learning Engineer @IQVIA.
Grad student at BUET.
Love Reinforcement Learning, Optimization, Logic & Chess.

[Falguni Das Shuvo @shuvoxcd01]

Congratulations to team Spain 🇪🇸
Spectacular football!
#FIFA2026

[Andrej Karpathy @karpathy]

Software horror: litellm PyPI supply chain attack.

Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords.

LiteLLM itself has 97 million downloads per month which is already terrible, but much worse, the contagion spreads to any project that depends on litellm. For example, if you did `pip install dspy` (which depended on litellm>=1.64.0), you'd also be pwnd. Same for any other large project that depended on litellm.

Afaict the poisoned version was up for only less than ~1 hour. The attack had a bug which led to its discovery - Callum McMahon was using an MCP plugin inside Cursor that pulled in litellm as a transitive dependency. When litellm 1.82.8 installed, their machine ran out of RAM and crashed. So if the attacker didn't vibe code this attack it could have been undetected for many days or weeks.

Supply chain attacks like this are basically the scariest thing imaginable in modern software. Every time you install any depedency you could be pulling in a poisoned package anywhere deep inside its entire depedency tree. This is especially risky with large projects that might have lots and lots of dependencies. The credentials that do get stolen in each attack can then be used to take over more accounts and compromise more packages.

Classical software engineering would have you believe that dependencies are good (we're building pyramids from bricks), but imo this has to be re-evaluated, and it's why I've been so growingly averse to them, preferring to use LLMs to "yoink" functionality when it's simple enough and possible.

[Falguni Das Shuvo @shuvoxcd01]

Blimey!

[Falguni Das Shuvo @shuvoxcd01]

@tahins 😁

[Falguni Das Shuvo @shuvoxcd01]

@GMJuditPolgar 8 for sure.

[Falguni Das Shuvo @shuvoxcd01]

@Sensodyne_US Hi @Sensodyne_US , Thanks for the reply. I've already contacted [email protected]. They replied that Haleon does not officially sell this product in Bangladesh. They have reached out to the appropriate department in France. I'll keep you posted if I get a reply.

[Falguni Das Shuvo @shuvoxcd01]

Is it even a legitimate product?
It says "made in USA" on the back.

@Sensodyne_US

[Falguni Das Shuvo @shuvoxcd01]

I didn't find any similar-looking products on the official Sensodyne websites. I'll be happy to provide more details if required.

[Falguni Das Shuvo @shuvoxcd01]

@GithubProjects Inspired!!🤟🤟

[Dieter Büchler @dtrbchlr]

@AmiiThinks @MPI_IS If you are interested in working with me at *the* RL powerhouse @UAlberta on robot learning on physical robots, please drop me a message.

Retweets welcome 🙏

[Falguni Das Shuvo @shuvoxcd01]

@dtrbchlr @AmiiThinks @MPI_IS @UAlberta I've always been fascinated by robot intelligence. RL + Robotics is my dream. I study RL, publish implementations, and try to fit NNs in ESP32 :). I'm taking my first steps alone due to a lack of collaboration. This opportunity would be like a dream come true. Super enthusiastic!

[Falguni Das Shuvo @shuvoxcd01]

Yep, it does struggle with watch faces!!!

[Falguni Das Shuvo @shuvoxcd01]

Happy New Year 🥳

[Falguni Das Shuvo @shuvoxcd01]

Merry Christmas !!

[Falguni Das Shuvo @shuvoxcd01]

Congratulations to @DGukesh for being the youngest world champion.

[Falguni Das Shuvo @shuvoxcd01]

📢 GridMind: A hands-on RL companion!

🎯 Compatible with gymnasium environments.
📚 Accompanies readers of Reinforcement Learning: An Introduction (2nd ed.).
🐍 Try it: pip install gridmind
🐙 Explore: https://t.co/sG0XQ24Azo

#ReinforcementLearning #GridMind

[Falguni Das Shuvo @shuvoxcd01]

🔗 GitHub: https://t.co/etmC74l18k

[Falguni Das Shuvo @shuvoxcd01]

🚶‍♂️🎉 Presenting: Random Walk environment! 🌟

Dive into this custom Gymnasium environment that simulates a Markov Reward Process. Perfect for hands-on exploration of RL concepts and making learning fun!

🔗 PyPI: https://t.co/Npz2sT7Dkj

#ReinforcementLearning #RandomWalkEnv #AI

[Falguni Das Shuvo @shuvoxcd01]

🚀 Implemented Tabular TD(0) Prediction! 🎉 It’s a small step in #ReinforcementLearning but a huge one for me. After digging into Sutton & Barto’s work, it feels great to see it in action.

Check it out here: https://t.co/lvWCyE688O 💻

#RL #MachineLearning #AI #OpenSource

[Google DeepMind @GoogleDeepMind]

Huge congratulations to @DemisHassabis and John Jumper on being awarded the 2024 Nobel Prize in Chemistry for protein structure prediction with #AlphaFold, along with David Baker for computational protein design.

This is a monumental achievement for AI, for computational biology, and science itself. 🧬