Tag
This paper identifies two critical gaps—the vocabulary gap and the verifier gap—that prevent current AI systems from achieving truly open-ended intelligence, and proposes a framework for evaluating and advancing innovation beyond fixed representational frames.
This paper identifies a vocabulary gap as the root cause why advanced encoders like ModernBERT underperform in learned sparse retrieval, and proposes Vocabulary Transfer (VT), a model-agnostic framework that migrates encoders to sparse-friendly vocabularies, achieving state-of-the-art on the BEIR benchmark.