aiqbench analysis
This study addresses the critical gap in NVIDIA's Nemotron retrieval models for Modern Greek, demonstrating how domain-specific adaptation via corpus mining and synthetic supervision can significantly boost RAG performance in legal, medical, and financial sectors. By fine-tuning embedders and LoRA-adapting readers on a new benchmark called HERA, researchers achieved substantial gains in answer correctness and citation faithfulness over both unadapted models and BM25 baselines. The release of these specialized tools provides an essential foundation for building reliable AI systems across specialist Greek-language applications.
Teaching Nemotron Greek: Mining a Corpus, Adapting Retrieval, and Grounding Generation for Modern Greek across Specialist Domains
arXiv:2608.05138v1 Announce Type: cross Abstract: Modern Greek is absent from NVIDIA's Nemotron retrieval models and from major multilingual retrieval benchmarks, despite being important for retrieval-augmented generation (RAG) in legal, energy, financial, and medical applications. We present an end-to-end adaptation of the Nemotron retrieval stack for Modern Greek, including corpus mining, synthetic supervision, retrieval model training, reranker adaptation, reader fine-tuning, and a new benchmark called HERA. Our study shows that a parameter-free BM25 baseline outperforms several off-the-shelf multilingual dense retrieval models on specialist Greek corpora. After fine-tuning on 65,773 Greek retrieval pairs, a Nemotron 1B embedder improves nDCG@10 from 0.362 to 0.835 and substantially outperforms its unadapted counterpart. The learned language competence transfers to general-domain Greek, although the advantage over BM25 remains domain-dependent. We further adapt a cross-encoder reranker and demonstrate consistent improvements across specialist domains. Finally, we LoRA-tune a Nemotron 30B-A3B mixture-of-experts reader for grounded generation, increasing judged answer correctness from 29.4% to 66.9% while significantly improving faithfulness and citation quality. We also introduce HERA, the first large-scale Greek benchmark for retrieval-augmented generation, and release our adapted models and benchmark to support future research on Greek-language RAG systems.
Abstract:Modern Greek is absent from NVIDIA's Nemotron retrieval models and from major multilingual retrieval benchmarks, despite being important for retrieval-augmented generation (RAG) in legal, energy, financial, and medical applications. We present an end-to-end adaptation of the Nemotron retrieval stack for Modern Greek, including corpus mining, synthetic supervision, retrieval model training, reranker adaptation, reader fine-tuning, and a new benchmark called HERA. Our study shows that a parameter-free BM25 baseline outperforms several off-the-shelf multilingual dense retrieval models on specialist Greek corpora. After fine-tuning on 65,773 Greek retrieval pairs, a Nemotron 1B embedder improves nDCG@10 from 0.362 to 0.835 and substantially outperforms its unadapted counterpart. The learned language competence transfers to general-domain Greek, although the advantage over BM25 remains domain-dependent. We further adapt a cross-encoder reranker and demonstrate consistent improvements across specialist domains. Finally, we LoRA-tune a Nemotron 30B-A3B mixture-of-experts reader for grounded generation, increasing judged answer correctness from 29.4% to 66.9% while significantly improving faithfulness and citation quality. We also introduce HERA, the first large-scale Greek benchmark for retrieval-augmented generation, and release our adapted models and benchmark to support future research on Greek-language RAG systems.
15 pages, 10 figures, 7 tables. Includes release of the HERA benchmark and Sophea Nemo RAG models
Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2608.05138 [eess.AS]
(or arXiv:2608.05138v1 [eess.AS] for this version)
https://doi.org/10.48550/arXiv.2608.05138 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 · 303 words
Related coverage
Article: Your Next DSL Author is a Language Model
InfoQ · Sep 16, 2026
Beyond the Chatbot: Co-Learning and Co-Teaching through a Dual-Persona Generative-AI Assistant
arXiv · Aug 27, 2026
Mastering Agentic Techniques: AI Agent Customization
NVIDIA · May 20, 2026
Evoflux: Inference-Time Evolution of Executable Tool Workflows for Compact Agents
arXiv · Jun 12, 2026