Rule library for the eleusis
inductive-reasoning environment. Each rule is a Python predicate over
(card, mainline).
test (26 rules): the rules of the
HF eleusis benchmark,
verified behaviorally identical to the benchmark repository's library.
train (52 rules): a training set matched to the test split — exactly 2x
its family mix and closely matched acceptance-rate distribution — with zero
behavioral overlap with the test rules (verified by extension-signature… See the full description on the dataset page:
https://huggingface.co/datasets/nph4rd/eleusis-hf-rules.
