# Information Aggregation for Multi-Head Attention with Routing-by-Agreement

arXiv: https://arxiv.org/abs/1904.03100 · HTML: https://mlsanity.com/paper/1904.03100/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Apr 2019.

## Built on by (most cited first, 13 of 13)

- [A Survey of Transformers](https://mlsanity.com/paper/2106.04554/) (2021)
- [Attention, please! A survey of Neural Attention Models in Deep Learning](https://mlsanity.com/paper/2103.16775/) (2021)
- [DSV: Exploiting Dynamic Sparsity to Accelerate Large-Scale Video DiT Training](https://mlsanity.com/paper/2502.07590/) (2025)
- [Improving Transformers with Dynamically Composable Multi-Head Attention](https://mlsanity.com/paper/2405.08553/) (2024)
- [ChatGPT-Like Large-Scale Foundation Models for Prognostics and Health Management: A Survey and Roadmaps](https://mlsanity.com/paper/2305.06472/) (2023)
- [FACTUM: Mechanistic Detection of Citation Hallucination in Long-Form RAG](https://mlsanity.com/paper/2601.05866/) (2026)
- [Devil is in the Uniformity: Exploring Diverse Learners within Transformer for Image Restoration](https://mlsanity.com/paper/2503.20174/) (2025)
- [Discrete Auto-regressive Variational Attention Models for Text Modeling](https://mlsanity.com/paper/2106.08571/) (2021)
- [Leveraging Advantages of Interactive and Non-Interactive Models for Vector-Based Cross-Lingual Information Retrieval](https://mlsanity.com/paper/2111.01992/) (2021)
- [EIT: Enhanced Interactive Transformer](https://mlsanity.com/paper/2212.10197/) (2022)
- [Finding the Pillars of Strength for Multi-Head Attention](https://mlsanity.com/paper/2305.14380/) (2023)
- [SOMP: Scalable Gradient Inversion for Large Language Models via Subspace-Guided Orthogonal Matching Pursuit](https://mlsanity.com/paper/2603.16761/) (2026)
- [Transformer-Enhanced Reinforcement Learning: Fundamentals and Applications in Communication Networks](https://mlsanity.com/paper/2606.05208/) (2026)

## Abstract (arXiv)

Multi-head attention is appealing for its ability to jointly extract different types of information from multiple representation subspaces. Concerning the information aggregation, a common practice is to use a concatenation followed by a linear transformation, which may not fully exploit the expressiveness of multi-head attention. In this work, we propose to improve the information aggregation for multi-head attention with a more powerful routing-by-agreement algorithm. Specifically, the routing algorithm iteratively updates the proportion of how much a part (i.e. the distinct information learned from a specific subspace) should be assigned to a whole (i.e. the final output representation), based on the agreement between parts and wholes. Experimental results on linguistic probing tasks and machine translation tasks prove the superiority of the advanced information aggregation over the standard linear transformation.