AI RESEARCH
LLMForge: Multi-Backend Hardware-Aware Neural Architecture Search with Infinite-Head Attention for Edge Language Models
arXiv CS.LG
•
ArXi:2605.17653v1 Announce Type: new Sub-billion-parameter Transformer language models are increasingly deployed on edge devices, where the privacy, latency, and operating-cost advantages of on-device inference are constrained by tight memory-bandwidth, energy, and thermal budgets that make architectural choice and accelerator-specific cost central to efficient inference.