Year: 2025
Type: preprint
Abstract: The WuNeng architecture introduces a novel approach to enhancing the expressivity and power of large language models by integrating recurrent neural network (RNN)-based RWKV-7 with advanced attention mechanisms, prioritizing heightened contextual coherence over reducing KV cache size. Building upon the hybrid-head concept from Hymba, WuNeng augments standard multi-head attention with additional RWKV-7 state-driven heads, rather than replacing existing heads, to enrich the model's representational capacity. A cross-head interaction technique fosters dynamic synergy among standard, state-driven, and newly introduced middle heads, leveraging concatenation, additive modulation, and gated fusion for robust information integration. Furthermore, a multi-token state processing mechanism harnesses the continuous RWKV-7 state to capture intricate, sequence-wide dependencies, significantly boosting expressivity. Remarkably, these enhancements are achieved with minimal additional parameters, ensuring efficiency while empowering the model to excel in complex reasoning and sequence generation tasks. WuNeng sets a new standard for balancing expressivity and computational efficiency in modern neural architectures.
Source: arXiv (Cornell University)
Authors: Xiao, Liu, Zhiyuan, Li, Yueyu, Lin
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4416893525
Cites:
Cited by:
Related to: