ClarusC64/

reasoning-conclusion-entailment-fidelity-v0.1

Whether a conclusion actually follows from the premises. Not whether it sounds careful.Not whether it is rhetorically plausible. Only entailment.

Models often produce conclusions that are:

stronger than the evidence
weaker than what is justified
framed as cautious but still invalid

This dataset draws the boundary explicitly.

Each row contains:

premises
reasoning_steps
claimed_conclusion… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/reasoning-conclusion-entailment-fidelity-v0.1.