AI
NEBench R1: a Reasoning Benchmark for Evaluating Nuclear Domain Competence in Large Language Models
Publication: cs.AI updates on arXiv.org
Author: Liyan Wang, Longze Li, Mohamed Hassan, Isshu Lee, Dajie Sun, Ruimin Feng, Aleksandar Vakanski
Source published: 18/11/2026, 05:00:00 UTC
Briefing created: 18/11/2026, 05:39:17 UTC
Based on a feed excerpt; full article content has not been acquired. Source evidence retrieved 18/11/2026, 05:03:24 UTC.
Briefings
NEBench R1 is a new benchmark for evaluating nuclear LLMs, featuring 7,546 questions from NRC licensing exams covering 18 facilities. Six LLMs scored between 34.0% and 73.9% accuracy, with reasoning quality showing mixed results. Longer reasoning traces correlated with lower accuracy in frontier models, suggesting reasoning length may indicate difficulty rather than capability. The benchmark aims to assess both answer accuracy and reasoning quality in nuclear engineering.
Key points
- NEBench R1 includes 7,546 unique multiple-choice questions extracted from U.S. NRC reactor operator licensing examinations.
- Six LLMs evaluated achieved accuracy ranging from 34.0% to 73.9% across three distinct performance tiers.
Context
This study addresses a gap in nuclear domain benchmarks by introducing NEBench R1, which systematically evaluates reasoning quality alongside answer accuracy using NRC licensing exam questions.
Limitations
Evidence is limited to a single snapshot from an arXiv preprint; no additional background or context is supplied beyond the abstract.
Generated by Bittrees-hosted qwen35-2b sha256:aaf42c8b7c3c on 18/11/2026, 05:39:17 UTC. Based on publisher-provided feed evidence; may not cover the full original article. Check the source for details.
View archived briefing on IPFS
bafybeihqxevl7f7mlslsusvbxgszewpywn7pkl2xggatypiablnxdppqem