Automatic Expert Discovery in LLM Upcycling via Sparse Interpolated Mixture-of-Experts
Chen, Shengzhuang, Wei, Ying, Schwarz, Jonathan Richard
–arXiv.org Artificial Intelligence
We present Sparse Interpolated Mixture-of-Experts (SIMoE) instruction-tuning, an end-to-end algorithm designed to fine-tune a dense pre-trained Large Language Model (LLM) into a MoE-style model that possesses capabilities in multiple specialized domains. During instruction-tuning, SIMoE automatically identifies multiple specialized experts under a specified sparsity constraint, with each expert representing a structurally sparse subset of the seed LLM's parameters that correspond to domain-specific knowledge within the data. SIMoE simultaneously learns an input-dependent expert merging strategy via a router network, leveraging rich cross-expert knowledge for superior downstream generalization that surpasses existing baselines. Empirically, SIMoE consistently achieves state-of-the-art performance on common instruction-tuning benchmarks while maintaining an optimal performance-compute trade-off compared to all baselines.
arXiv.org Artificial Intelligence
Jun-17-2025
- Country:
- Asia > Middle East > UAE (0.46)
- Genre:
- Research Report (1.00)
- Industry:
- Law > Litigation (0.40)
- Technology: