Qwen-3-4B

The 'Qwen-3-4B' dataset/benchmark is used to evaluate the performance and efficiency of deploying large language models (LLMs) on resource-constrained spatial neural processing units (NPUs).

Papers using Qwen-3-4B (2)

From Human Guidance to Autonomy: Agent Skill System for End-to-End LLM Deployment on Spatial NPUs2026Implicit Hierarchical GRPO: Decoupling Tool Invocation from Execution for Tool-Integrated Mathematical Reasoning2026