Blog

2026

Curiosity-Critic: Rewarding Cumulative Progress

11 minute read

Published: April 21, 2026

Tags: 2026, Reinforcement Learning, Research, World Models

Agent trajectories at step 30,000 on a stochastic grid world. — Seven methods explore a 30x30 grid world guided by curiosity. The left half (green) is learnable; the right half (grey) is pure noise. [Live animation: youtu.be/Jv1n346TWbQ, Code: github.com/vinbhaskara/Curiosity-Critic, Paper: arxiv.org/abs/2604.18701]

2023

Hosting Jupyter Notebooks on Slurm

2 minute read

Published: November 11, 2023

Tags: 2023, MLOps

Photo generated with AI by Bing. — Slurm is bliss. Jupyter is a blessing. But Jupyter on Slurm? That is where the trouble begins. This guide serves as a bookmark for the exact steps needed to tunnel a Jupyter server through Slurm without the headache.