datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
kepler-arc-agi-3-traces
Kepler 1.0 ARC-AGI-3 trace corpus
Run artifacts from Kepler 1.0, an open-source agent harness for the 25 public
ARC-AGI-3 games. A stock CLI coding agent
encodes its theory of each game as an executable world_model.py, certifies it
against the full recorded interaction history, plans inside the certified
model, and acts through a guarded channel that voids the plan on the first
misprediction.
Project page ·
Code ·
Paper ·
Integrity record
The canonical release contains two… See the full description on the dataset page: https://huggingface.co/datasets/cveinnt/kepler-arc-agi-3-traces.Kepler-bench
Kepler astro-bench v0.1
The benchmark behind Orionfold/Kepler-GGUF —
a verifier-checked set of astrodynamics and quantitative-astrophysics word problems, each with a
single numeric gold answer and a programmatic verifier that doubles as a reinforcement-learning
reward.
What's here
File
Rows
Purpose
pool.jsonl
120
Training / selection pool — 16 formula families (9 orbital, 7 astrophysics), 3 difficulty tiers.
heldout.jsonl
44
External curveball… See the full description on the dataset page: https://huggingface.co/datasets/Orionfold/Kepler-bench.