SmolDataEnvs launches: 5k+ verifiable RL tasks on real Kaggle data, fully open
SergioPaniego · x · 2026-09-28
SmolDataEnvs is out: 5,000+ verifiable RL tasks for hill-climbing small models in code and data science, led by @adithyask. Positioned as the next step beyond Wordle and GSM8K, built on real Kaggle data. Fully open — environments, SFT trajectories, and notebooks for SFT and RL are all released.
Related event: SmolDataEnvs Releases 5,500+ Verifiable RL Tasks on Real Kaggle Data(2 posts)→
More from Research
- SAGE Uses Topological Guidance to Fix Long-Horizon LLM Reasoning Biases — Xinyue Zeng · 2026-09-28
- LightMIS: 0.13M-Param Medical Segmentation Net Cuts 99% Params, Matches Accuracy — Andrei Arhire · 2026-09-28
- Eval tips for automated optimizers: use AUC over accuracy, sample 10x and check consensus to cut hallucinations — a_karvonen · 2026-09-28
- Three Simple Inference Tricks Significantly Improve Activation Oracle Results — a_karvonen · 2026-09-28
- Paper: Classifier Judge Matches LLM Judges and Copies 96% of Their Errors — deliprao · 2026-09-28
- Schmidhuber's 2008 'Compression Progress' paper: curiosity is just an algorithm — anselm · 2026-09-28