RAC: Rectified Flow Auto Coder

2026-03-06

AI Summary

This work addresses the inherent trade-off between reconstruction fidelity and generative capability in conventional variational autoencoders (VAEs), along with their high computational cost and parameter inefficiency. The authors propose a novel autoencoder architecture that, for the first time, integrates Rectified Flow into the autoencoding framework. By employing a multi-step rectifiable decoding process for progressive refinement and leveraging a time-reversal mechanism to enable shared encoder–decoder dynamics, the model achieves bidirectional inference. This unification of encoding and decoding pathways substantially reduces both computational overhead and model parameters. Experimental results demonstrate that the proposed method outperforms state-of-the-art VAE approaches in both reconstruction and generation tasks, achieving approximately 70% lower computational cost and nearly 41% fewer parameters.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

In this paper, we propose a Rectified Flow Auto Coder (RAC) inspired by Rectified Flow to replace the traditional VAE: 1. It achieves multi-step decoding by applying the decoder to flow timesteps. Its decoding path is straight and correctable, enabling step-by-step refinement. 2. The model inherently supports bidirectional inference, where the decoder serves as the encoder through time reversal (hence Coder rather than encoder or decoder), reducing parameter count by nearly 41%. 3. This generative decoding method improves generation quality since the model can correct latent variables along the path, partially addressing the reconstruction--generation gap. Experiments show that RAC surpasses SOTA VAEs in both reconstruction and generation with approximately 70% lower computational cost.

Problem

• Variational Autoencoder
• Reconstruction-Generation Gap
• Bidirectional Inference
• Multi-step Decoding
• Computational Efficiency

Innovation

• Rectified Flow
• Auto Coder
• Bidirectional Inference
• Multi-step Decoding
• Latent Correction