{"id":"arxiv/2602.10544","url":"https://arxiv.org/abs/2602.10544","tags":["cs.LG","math.NA"],"score":0,"title":"Bridging the Compression-Precision Paradox: A Hybrid Architecture for Clinical EEG Report Generation with Guaranteed Measurement Accuracy","domain":"cs.LG","source":"arxiv","authors":["Wuyang Zhang","Zhen Luo","Chuqiao Gu","Jianming Ma","Yebo Cao","Wangming Yuan","Yinzhi Jin"],"reviews":0,"subject":"cs.LG","summary":"Automated EEG monitoring requires clinician-level precision for seizure detection and reporting. Clinical EEG recordings exceed LLM context windows, requiring extreme compression (400:1+ ratios) that destroys fine-grained temporal precision. A 0.5 Hz error distinguishes absence epilepsy from Lennox-Gastaut syndrome. LLMs lack inherent time-series comprehension and rely on statistical associations from compressed representations. This dual limitation causes systems to hallucinate clinically incorrect measurement values. We separate measurement extraction from text generation. Our hybrid…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-02-11","reactions":{}}