144 human-validated sentence pairs from the es_wikibooks source of ESNLIR, used only as the
pool to draw few-shot examples from in An Analysis of the Performance of Large Language Models in
Spanish NLI Datasets with Causal Relationships (IBERAMIA 2026, to appear). Part of the
ESNLIR-LLM collection, used in
Pacolas/NLI-via-LLM.
Validated by two annotators; only pairs where both assigned the same label as the dataset were kept.… See the full description on the dataset page:
https://huggingface.co/datasets/Flaglab/esnlir-fewshot-pool.
