DeepSeek

DeepSeek is a benchmark dataset used to evaluate the performance of fine-grained Mixture-of-Experts models in terms of expert reuse and efficiency during inference.

Papers using DeepSeek (2)

When Can You Debias an LLM Judge? Identifiability Limits, a Test, and Designs for Top-k Ranking2026ReMoE: Boosting Expert Reuse through Router Fine-Tuning in Memory-Constrained MoE LLM Inference2026