aiqbench analysis
This paper introduces NE-R1, a framework that optimizes Named Entity Recognition by combining multi-task instruction tuning with reinforcement learning to manage retrieval-augmented generation more efficiently. By implementing a "retrieval-on-demand" mechanism and a multi-dimensional reward system, the model reduces noise from external knowledge while improving accuracy on long-tail entities. The approach demonstrates significant performance gains in both in-domain and zero-shot cross-domain evaluations, addressing key limitations of current LLM-based NER systems.
NE-R1: Enhancing Named Entity Recognition Model via Reinforcement Learning
arXiv:2609.02366v1 Announce Type: new Abstract: Named Entity Recognition (NER) has achieved substantial progress since the advent of large language models (LLMs). Nevertheless, the recognition of long-tail and domain-specific entities remains challenging due to the deficiency in parametric knowledge. Retrieval-augmented generation (RAG) offers a promising remedy by injecting external knowledge, but it also introduces noise and unnecessary cost when dealing with familiar cases. In this paper, we propose NE-R1, a novel framework for adaptive retrieval-augmented NER. We design a "retrieval-on-demand" mechanism for NER. Then we integrate it into models by a two-stage training method: (1) multi-task instruction tuning initialization; (2) end-to-end RL optimization with CoT. To achieve reasonable selection between parameterized and external knowledge, we design a multi-dimensional reward considering both accuracy and retrieval benefit. NE-R1 achieves state-of-the-art performance on various benchmarks, with an average F1 score gain of 2.52% in in-domain evaluation and 1.18% in zero-shot cross-domain evaluation.
Abstract:Named Entity Recognition (NER) has achieved substantial progress since the advent of large language models (LLMs). Nevertheless, the recognition of long-tail and domain-specific entities remains challenging due to the deficiency in parametric knowledge. Retrieval-augmented generation (RAG) offers a promising remedy by injecting external knowledge, but it also introduces noise and unnecessary cost when dealing with familiar cases. In this paper, we propose NE-R1, a novel framework for adaptive retrieval-augmented NER. We design a "retrieval-on-demand" mechanism for NER. Then we integrate it into models by a two-stage training method: (1) multi-task instruction tuning initialization; (2) end-to-end RL optimization with CoT. To achieve reasonable selection between parameterized and external knowledge, we design a multi-dimensional reward considering both accuracy and retrieval benefit. NE-R1 achieves state-of-the-art performance on various benchmarks, with an average F1 score gain of 2.52% in in-domain evaluation and 1.18% in zero-shot cross-domain evaluation.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2609.02366 [cs.CL]
(or arXiv:2609.02366v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2609.02366 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 241 words
Related coverage
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · Sep 24, 2026
Enhancing Scientific Named Entity Recognition via Large Language Models: A Type-driven Multi-task Learning Approach
arXiv · Aug 18, 2026
LA-RL: Label-Aware Self-Reflection for Reinforcement Learning in Information Extraction
arXiv · Jul 28, 2026
FSE: Continual Learning for Named Entity Recognition by Fast-Slow Experts
arXiv · Jul 28, 2026