New framework enhances MoE LLMs on noisy analog hardware

Researchers have introduced ROMER, a post-training calibration framework designed to enhance the robustness of Mixture-of-Experts (MoE) Large Language Models (LLMs) when deployed on analog Compute-in-Memory (CIM) systems. This framework addresses hardware imperfections in CIM by replacing underutilized experts and recalibrating router decisions to maintain load balance and optimal routing under noisy conditions. Experiments show ROMER significantly reduces perplexity for models like DeepSeek-MoE, Qwen-MoE, and OLMoE when subjected to real-chip noise. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Improves the viability of deploying LLMs on energy-efficient analog hardware by mitigating noise-induced performance degradation.

RANK_REASON The cluster contains an academic paper detailing a new method for improving LLM performance on specific hardware. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• ROMER
• LLMs
• MoE
• CIM
• DeepSeek-MoE
• Qwen-MoE
• OLMoE

• paper
• infra