Ayush Das
11 posts
Ayush Das
@ayushdas45
Training models & myself—both overfit sometimes 🤖💥
Katılım Şubat 2025
30 Takip Edilen4 Takipçiler
I can point you to at least 5 jobs right now in Germany if you understand:
1. Linear Regression
2. Time Series Forecasting (ARIMA)
3. Bayes’ theorem
4. You can write code in Python
5. Add ons: Logistic Regression, Decision Trees, Clustering algorithms and Monte Carlo Simulation
Dr. Kalle@TAnalysen
@mayukh_panja I wanted to work as a statistician. But 1) there are no jobs as such on the market 2) I think I keep being rejected bc I did not study statistics per se, but I had it in my studies and later of course in my PhD (where I took 4 courses on it again and all the math didnt change).
English
💫 Introducing RLP: Reinforcement Learning Pretraining—information-driven, verifier-free objective that teaches models to think before they predict
🔥+19% vs BASE on Qwen3-1.7B
🚀+35% vs BASE on Nemotron-Nano-12B
📄Paper: github.com/NVlabs/RLP/blo…
📝Blog: research.nvidia.com/labs/adlr/RLP/
English
🌠Be Doraemon — the Real One ✨
>🚪 Comes from the future with zero agenda
>🙅♂️ Never asks for your data, never sells your dreams
>🎒 Pulls exactly the gadget you need — not the one that tracks you
>🧸 Fixes Nobita’s mess without judging, without ads, without “premium pocket”
>📜 Teaches that real help doesn’t come with terms of service
>💫 Blue, round, pure… and forever free.
--------------------------------------------
If they tried to monetize one Doraemon gadget — which would you protect with your life?
👉 Anywhere Door? 🚪
👉 Time Machine? ⏳
👉 Take-copter? 🪂
👉 Or the thing no one sells: his kindness? ❤️
English
🔥 PYTHON vs C++ DSA BATTLE
LeetCode Slayer or AI/ML Monster? 🔥
Python's my jam—data projects galore, syntax like poetry!
But C++ tempts w/ raw speed & low-level control... 😈
Both, or Nah (Rust!)? roast my dilemma!
#DSA #PythonVsCpp #AI #MachineLearning
English
