{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,9]],"date-time":"2026-04-09T14:37:30Z","timestamp":1775745450743,"version":"3.50.1"},"reference-count":38,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE J. Sel. Areas Commun."],"published-print":{"date-parts":[[2026]]},"DOI":"10.1109\/jsac.2025.3641905","type":"journal-article","created":{"date-parts":[[2025,12,9]],"date-time":"2025-12-09T18:33:28Z","timestamp":1765305208000},"page":"2365-2377","source":"Crossref","is-referenced-by-count":1,"title":["Go Gentle Into the Final Dance: A Benchmark for Evaluating LLMs in Telecom"],"prefix":"10.1109","volume":"44","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8311-6601","authenticated-orcid":false,"given":"Yushen","family":"Lin","sequence":"first","affiliation":[{"name":"Department of Electrical and Electronic Engineering, The University of Manchester, Manchester, U.K."}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5280-384X","authenticated-orcid":false,"given":"Zhiguo","family":"Ding","sequence":"additional","affiliation":[{"name":"Department of Electrical and Electronic Engineering, The University of Manchester, Manchester, U.K."}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6859-3645","authenticated-orcid":false,"given":"Ruichen","family":"Zhang","sequence":"additional","affiliation":[{"name":"College of Computing and Data Science, Nanyang Technological University, Jurong West, Singapore"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7642-1755","authenticated-orcid":false,"given":"Daniel K. C.","family":"So","sequence":"additional","affiliation":[{"name":"Department of Electrical and Electronic Engineering, The University of Manchester, Manchester, U.K."}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Measuring massive multitask language understanding","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Hendrycks"},{"key":"ref2","article-title":"Training verifiers to solve math word problems","author":"Cobbe","year":"2021","journal-title":"arXiv:2110.14168"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-acl.824"},{"key":"ref4","article-title":"Tag-LLM: Repurposing general-purpose LLMs for specialized domains","author":"Shen","year":"2024","journal-title":"arXiv:2402.05140"},{"key":"ref5","article-title":"Uncertainty-aware hybrid inference with on-device small and remote large language models","author":"Oh","year":"2024","journal-title":"arXiv:2412.12687"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/MNET.2025.3541078"},{"key":"ref7","article-title":"TeleQnA: A benchmark dataset to assess large language models telecommunications knowledge","author":"Maatouk","year":"2023","journal-title":"arXiv:2310.15051"},{"key":"ref8","article-title":"Wireless hallucination in generative AI-enabled communications: Concepts, issues, and solutions","author":"Wang","year":"2025","journal-title":"arXiv:2503.06149"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.23919\/JCIN.2024.10582827"},{"key":"ref10","article-title":"MMLU-Pro: A more robust and challenging multi-task language understanding benchmark","author":"Wang","year":"2024","journal-title":"arXiv:2406.01574"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/tcomm.2025.3597815"},{"key":"ref12","article-title":"FEABench: Evaluating language models on multiphysics reasoning ability","author":"Mudur","year":"2025","journal-title":"arXiv:2504.06260"},{"key":"ref13","article-title":"Qwen2.5\u20131M technical report","author":"Yang","year":"2025","journal-title":"arXiv:2501.15383"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.acl-long.1285"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/MCOM.001.2300473"},{"key":"ref16","first-page":"609","article-title":"TelBench: A benchmark for evaluating telco-specific large language models","volume-title":"Proc. Conf. Empirical Methods Natural Lang. Processing: Ind. Track","author":"Lee"},{"key":"ref17","article-title":"Bidirectional encoder representations from transformers (BERT) for question answering in the telecom domain: Adapting a BERT-like language model to the telecom domain using the ELECTRA pre-training approach","author":"Holm","year":"2021"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/d19-1260"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/COMSNETS59351.2024.10427044"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/ICMLCN59089.2024.10624786"},{"key":"ref21","first-page":"20","article-title":"SPEC5G: A dataset for 5G cellular network protocol analysis","volume-title":"Proc. Findings Assoc. Comput. Linguistics, IJCNLP-AACL","author":"Karim"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/lwc.2017.2759114"},{"key":"ref23","article-title":"The llama 3 herd of models","author":"Grattafiori","year":"2024","journal-title":"arXiv:2407.21783"},{"key":"ref24","article-title":"Quiet-STaR: Language models can teach themselves to think before speaking","author":"Zelikman","year":"2024","journal-title":"arXiv:2403.09629"},{"key":"ref25","article-title":"Automated design of agentic systems","author":"Hu","year":"2024","journal-title":"arXiv:2408.08435"},{"key":"ref26","article-title":"LLM multi-agent systems: Challenges and open problems","author":"Han","year":"2024","journal-title":"arXiv:2402.03578"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2024\/890"},{"key":"ref28","article-title":"AgentInstruct: Toward generative teaching with agentic flows","author":"Mitra","year":"2024","journal-title":"arXiv:2407.03502"},{"key":"ref29","article-title":"Agentic retrieval-augmented generation: A survey on agentic RAG","author":"Singh","year":"2025","journal-title":"arXiv:2501.09136"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/tbdata.2025.3618474"},{"key":"ref31","article-title":"Qwen2.5 technical report","author":"Hui","year":"2024","journal-title":"arXiv:2412.15115"},{"key":"ref32","volume-title":"QwQ-32B: Embracing the Power of Reinforcement Learning","year":"2025"},{"key":"ref33","article-title":"DeepSeek LLM: Scaling open-source language models with longtermism","author":"Bi","year":"2024","journal-title":"arXiv:2401.02954"},{"key":"ref34","article-title":"When more is less: Understanding chain-of-thought length in LLMs","volume-title":"Proc. Workshop Reasoning Planning Large Lang. Models (RP4LLM), Int. Conf. Learn. Represent. (ICLR)","author":"Wu"},{"key":"ref35","first-page":"5872","article-title":"Masked thought: Simply masking partial reasoning steps can improve mathematical reasoning learning of language models","volume-title":"Proc. 62nd Annu. Meeting Assoc. Comput. Linguistics (Volume 1: Long Papers)","author":"Chen"},{"key":"ref36","article-title":"Chain-of-thought prompting elicits reasoning in large language models","author":"Wei","year":"2022","journal-title":"arXiv:2201.11903"},{"key":"ref37","article-title":"Self-consistency improves chain of thought reasoning in language models","author":"Wang","year":"2022","journal-title":"arXiv:2203.11171"},{"key":"ref38","article-title":"ReAct: Synergizing reasoning and acting in language models","author":"Yao","year":"2022","journal-title":"arXiv:2210.03629"}],"container-title":["IEEE Journal on Selected Areas in Communications"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/49\/11372474\/11288986.pdf?arnumber=11288986","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T20:48:35Z","timestamp":1771879715000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11288986\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":38,"URL":"https:\/\/doi.org\/10.1109\/jsac.2025.3641905","relation":{},"ISSN":["1558-0008","0733-8716"],"issn-type":[{"value":"1558-0008","type":"electronic"},{"value":"0733-8716","type":"print"}],"subject":[],"published":{"date-parts":[[2026]]}}}