{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,27]],"date-time":"2026-05-27T06:05:30Z","timestamp":1779861930018,"version":"3.53.1"},"reference-count":57,"publisher":"IEEE","license":[{"start":{"date-parts":[[2026,4,26]],"date-time":"2026-04-26T00:00:00Z","timestamp":1777161600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,4,26]],"date-time":"2026-04-26T00:00:00Z","timestamp":1777161600000},"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":[],"published-print":{"date-parts":[[2026,4,26]]},"DOI":"10.1109\/ispass69572.2026.00013","type":"proceedings-article","created":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T19:39:19Z","timestamp":1779824359000},"page":"15-27","source":"Crossref","is-referenced-by-count":0,"title":["Characterizing State Space Model and Hybrid Language Model Performance with Long Context"],"prefix":"10.1109","author":[{"given":"Saptarshi","family":"Mitra","sequence":"first","affiliation":[{"name":"University of California, Irvine,Electrical Engineering and Computer Science,Irvine,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rachid","family":"Karami","sequence":"additional","affiliation":[{"name":"University of California, Irvine,Electrical Engineering and Computer Science,Irvine,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Haocheng","family":"Xu","sequence":"additional","affiliation":[{"name":"University of California, Irvine,Electrical Engineering and Computer Science,Irvine,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sitao","family":"Huang","sequence":"additional","affiliation":[{"name":"University of California, Irvine,Electrical Engineering and Computer Science,Irvine,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hyoukjun","family":"Kwon","sequence":"additional","affiliation":[{"name":"University of California, Irvine,Electrical Engineering and Computer Science,Irvine,USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Phi-3 technical report: A highly capable language model locally on your phone","volume-title":"arXiv preprint arXiv:2404.14219","author":"Abdin","year":"2024"},{"key":"ref2","article-title":"Gpt-4 technical report","volume-title":"arXiv preprint arXiv:2303.08774","author":"Achiam","year":"2023"},{"key":"ref3","first-page":"4895","article-title":"GQA: Training generalized multi-query transformer models from multi-head checkpoints","volume-title":"Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing","author":"Ainslie"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.acl-long.183"},{"key":"ref5","article-title":"Titans: Learning to memorize at test time","author":"Behrouz","year":"2024","journal-title":"arXiv preprint arXiv:2501.00663"},{"key":"ref6","article-title":"Longformer: The long-document transformer","author":"Beltagy","year":"2020","journal-title":"arXiv:2004.05150"},{"key":"ref7","article-title":"GPT-Neo: Large Scale Autoregressive Language Modeling with Mesh-Tensorflow","author":"Black","year":"2021"},{"key":"ref8","article-title":"Falcon-h1r: Pushing the reasoning frontiers with a hybrid model for efficient testtime scaling","author":"Chaabane","year":"2026","journal-title":"arXiv preprint arXiv:2601.02346"},{"key":"ref9","article-title":"Longlora: Efficient fine-tuning of long-context large language models","volume-title":"The Twelfth International Conference on Learning Representations (ICLR)","author":"Chen"},{"key":"ref10","article-title":"Quamba2: A robust and scalable post-training quantization framework for selective state space models","author":"Chiang","year":"2025","journal-title":"arXiv preprint arXiv:2503.22879"},{"key":"ref11","article-title":"Think you have solved question answering? try arc, the ai2 reasoning challenge","author":"Clark","year":"2018","journal-title":"arXiv preprint arXiv:1803.05457"},{"key":"ref12","first-page":"16 344","article-title":"Flashattention: Fast and memory-efficient exact attention with io-awareness","volume-title":"Proceedings of the Advances in Neural Information Processing Systems (NeurIPS)","volume":"35","author":"Dao"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1189"},{"key":"ref14","article-title":"Transformers are ssms: Generalized models and efficient algorithms through structured state space duality","author":"Dao","year":"2024","journal-title":"arXiv preprint arXiv:2405.21060"},{"key":"ref15","article-title":"Hymba: A Hybrid-head Architecture for Small Language Models","author":"Dong","year":"2024"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1145\/3721146.3721961"},{"key":"ref17","article-title":"Nemotron-flash: Towards latency-optimal hybrid small language models","author":"Fu","year":"2025","journal-title":"arXiv preprint arXiv:2511.18890"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW59228.2023.00491"},{"key":"ref19","article-title":"The zamba2 suite: Technical report","volume-title":"arXiv preprint arXiv:2411.15242","author":"Glorioso","year":"2024"},{"key":"ref20","article-title":"The llama 3 herd of models","author":"Grattafiori","year":"2024","journal-title":"arXiv preprint arXiv:2407.21783"},{"key":"ref21","article-title":"mamba","author":"Gu","year":"2023"},{"key":"ref22","article-title":"Mamba: Linear-time sequence modeling with selective state spaces","author":"Gu","year":"2024"},{"key":"ref23","article-title":"Efficiently modeling long sequences with structured state spaces","author":"Gu","year":"2022"},{"issue":"39","key":"ref24","first-page":"851","article-title":"Parallel prefix sum (scan) with cuda","volume-title":"GPU gems","volume":"3","author":"Harris"},{"key":"ref25","article-title":"Measuring massive multitask language understanding","volume-title":"International Conference on Learning Representations (ICLR)","author":"Hendrycks"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1145\/3725843.3756115"},{"key":"ref27","article-title":"Scaling laws for neural language models","volume":"abs\/2001.08361","author":"Kaplan","year":"2020","journal-title":"CoRR"},{"key":"ref28","article-title":"Nongemm bench: Understanding the performance horizon of the latest ml workloads with nongemm workloads","author":"Karami","year":"2025"},{"key":"ref29","first-page":"106 519","article-title":"Babilong: Testing the limits of llms with long context reasoning-in-a-haystack","volume-title":"Advances in Neural Information Processing Systems","volume":"37","author":"Kuratov"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1145\/3600006.3613165"},{"key":"ref31","first-page":"155","article-title":"{InfiniGen}: Efficient generative inference of large language models with dynamic {KV} cache management","volume-title":"18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24)","author":"Lee"},{"key":"ref32","first-page":"3214","article-title":"TruthfulQA: Measuring how models mimic human falsehoods","volume-title":"Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Lin"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0943"},{"key":"ref34","article-title":"Accelerating semantic image segmentation on fpga","author":"Mitra","year":"2021"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/IISWC63097.2024.00024"},{"key":"ref36","article-title":"Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models","year":"2025"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/iscc58397.2023.10217850"},{"key":"ref38","article-title":"Can mamba learn how to learn? a comparative study on in-context learning tasks","volume-title":"Proceedings of the 41st International Conference on Machine Learning, ser. ICML\u201924","author":"Park"},{"key":"ref39","article-title":"Deep learning inference in facebook data centers: Characterization, performance optimizations and hardware implications","author":"Park","year":"2018","journal-title":"arXiv preprint arXiv:1811.09886"},{"key":"ref40","article-title":"Pytorch: An imperative style, high-performance deep learning library","volume-title":"Proceedings of the Advances in Neural Information Processing Systems (NeurIPS)","volume":"32","author":"Paszke"},{"key":"ref41","article-title":"Efficiently scaling transformer inference","author":"Pope","year":"2022"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/HOTI59126.2023.00022"},{"issue":"8","key":"ref43","first-page":"9","article-title":"Language models are unsupervised multitask learners","volume":"1","author":"Radford","year":"2019","journal-title":"OpenAI blog"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1145\/3474381"},{"key":"ref45","first-page":"18 506","article-title":"MiniKV: Pushing the limits of 2-bit KV cache via compression and system co-design for efficient long context inference","author":"Sharma","year":"2025","journal-title":"Findings of the Association for Computational Linguistics: ACL 2025"},{"key":"ref46","article-title":"Flexgen: high-throughput generative inference of large language models with a single gpu","volume-title":"Proceedings of the 40th International Conference on Machine Learning, ser. ICML \u201923","author":"Sheng"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52734.2025.02435"},{"key":"ref48","article-title":"Llama: Open and efficient foundation language models","volume":"abs\/2302.13971","author":"Touvron","year":"2023","journal-title":"ArXiv"},{"key":"ref49","article-title":"Attention is all you need","volume-title":"Proceedings of the Advances in Neural Information Processing Systems (NeurIPS)","volume":"30","author":"Vaswani"},{"key":"ref50","article-title":"Attention is all you need","author":"Vaswani","year":"2023"},{"key":"ref51","article-title":"An Empirical Study of Mamba-based Language Models","author":"Waleffe","year":"2024"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-demos.6"},{"key":"ref53","article-title":"Qwen3 technical report","volume-title":"arXiv preprint arXiv:2505.09388","author":"Yang","year":"2025"},{"key":"ref54","first-page":"4791","article-title":"HellaSwag: Can a machine really finish your sentence?","volume-title":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","author":"Zellers"},{"key":"ref55","article-title":"Tinyllama: An open-source small language model","author":"Zhang","year":"2024","journal-title":"arXiv preprint arXiv:2401.02385"},{"key":"ref56","first-page":"15 262","article-title":"\u221eBench: Extending long context evaluation beyond 100K tokens","volume-title":"Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Zhang"},{"key":"ref57","article-title":"Falcon-h1: A family of hybrid-head language models redefining efficiency and performance","author":"Zuo","year":"2025","journal-title":"arXiv preprint arXiv:2507.22448"}],"event":{"name":"2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)","location":"Seoul, Korea, Republic of","start":{"date-parts":[[2026,4,26]]},"end":{"date-parts":[[2026,4,28]]}},"container-title":["2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11527204\/11527232\/11527267.pdf?arnumber=11527267","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,27]],"date-time":"2026-05-27T05:36:34Z","timestamp":1779860194000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11527267\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,26]]},"references-count":57,"URL":"https:\/\/doi.org\/10.1109\/ispass69572.2026.00013","relation":{},"subject":[],"published":{"date-parts":[[2026,4,26]]}}}