LawngNLI: A Long-Premise Benchmark for In-Domain Generalization from Short to Long Contexts and for Implication-Based Retrieval.

William Bruno, Dan Roth

Browse the full EMNLP paper archive.