Search papers, labs, and topics across Lattice.
Télécom Paris, Institut Polytechnique de Paris
3
0
6
Smaller language models can now match the factual accuracy of their larger counterparts by dynamically querying a knowledge base during generation.
Agentic harnesses can significantly enhance LLM performance on deontic reasoning tasks, but not without introducing risks of degradation in numerical accuracy.
Domain-agnostic entity linking is now possible out-of-the-box, thanks to an LLM-powered framework that needs no training data.
