Recently, several specialized instruction-tuned Large Language Models (LLMs) for Named Entity Recognition (NER) have emerged. Compared with traditional NER approaches, these models have demonstrated strong generalization capabilities. Existing LLMs primarily focus on addressing zero-shot NER on Out-of-Domain inputs, while fine-tuning on an extensive number of entity classes that often highly or completely overlap with test sets. In this work instead, we propose SLIMER, an approach designed to tackle never-seen-before entity tags by instructing the model on fewer examples, and by leveraging a prompt enriched with definition and guidelines. Experiments demonstrate that definition and guidelines yield better performance, faster and more robust learning, particularly when labelling unseen named entities. Furthermore, SLIMER performs comparably to state-of-the-art approaches in out-of-domain zero-shot NER, while being trained in a more fair, though certainly more challenging, setting.

The Single Frontier Bi-Directional Search (SBS) framework was recently introduced. A node in SBS corresponds to a pair of states, one from each of the frontiers and it uses front-to- front heuristics. In this paper we present an enhanced version of SBS, called eSBS, where pruning and caching techniques are applied, which significantly reduce both time and memory needs of SBS. We then present a hybrid of eSBS and IDA∗ which potentially uses only the square root of the memory required by A∗ but enables to prune many nodes that IDA∗ would generate. Experimental results show the benefit of our new approaches on a number of domains.