# Dual-Rerank: Fusing Causality and Utility for Industrial Generative Reranking

arXiv: https://arxiv.org/abs/2604.07420 · HTML: https://mlsanity.com/paper/2604.07420/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 35 references that are in the map.
- First posted on arXiv 8 Apr 2026.

## Builds on (most cited first, 22 of 22)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling](https://mlsanity.com/paper/1412.3555/) (2014)
- [Wide & Deep Learning for Recommender Systems](https://mlsanity.com/paper/1606.07792/) (2016)
- [Non-Autoregressive Neural Machine Translation](https://mlsanity.com/paper/1711.02281/) (2017)
- [Deep & Cross Network for Ad Click Predictions](https://mlsanity.com/paper/1708.05123/) (2017)
- [Learning a Deep Listwise Context Model for Ranking Refinement](https://mlsanity.com/paper/1804.05936/) (2018)
- [SetRank: Learning a Permutation-Invariant Ranking Model for Information Retrieval](https://mlsanity.com/paper/1912.05891/) (2019)
- [LLM4Rerank: LLM-based Auto-Reranking Framework for Recommendations](https://mlsanity.com/paper/2406.12433/) (2024)
- [Seq2Slate: Re-ranking and Slate Optimization with RNNs](https://mlsanity.com/paper/1810.02019/) (2018)
- [Beyond Greedy Ranking: Slate Optimization via List-CVAE](https://mlsanity.com/paper/1803.01682/) (2018)
- [Neural Re-ranking in Multi-stage Recommender Systems: A Review](https://mlsanity.com/paper/2202.06602/) (2022)
- [Non-autoregressive Generative Models for Reranking Recommendation](https://mlsanity.com/paper/2402.06871/) (2024)
- [Distilled Decoding 1: One-step Sampling of Image Auto-regressive Models with Flow Matching](https://mlsanity.com/paper/2412.17153/) (2024)
- [Revisit Recommender System in the Permutation Prospective](https://mlsanity.com/paper/2102.12057/) (2021)
- [GRN: Generative Rerank Network for Context-wise Recommendation](https://mlsanity.com/paper/2104.00860/) (2021)
- [Variation Control and Evaluation for Generative SlateRecommendations](https://mlsanity.com/paper/2102.13302/) (2021)
- [KuaiLive: A Real-time Interactive Dataset for Live Streaming Recommendation](https://mlsanity.com/paper/2508.05633/) (2025)
- [A Generative Re-ranking Model for List-level Multi-objective Optimization at Taobao](https://mlsanity.com/paper/2505.07197/) (2025)
- [Context-aware Reranking with Utility Maximization for Recommendation](https://mlsanity.com/paper/2110.09059/) (2021)
- [NLGR: Utilizing Neighbor Lists for Generative Rerank in Personalized Recommendation Systems](https://mlsanity.com/paper/2502.06097/) (2025)
- [Comprehensive List Generation for Multi-Generator Reranking](https://mlsanity.com/paper/2504.15625/) (2025)
- [DualGR: Generative Retrieval with Long and Short-Term Interests Modeling](https://mlsanity.com/paper/2511.12518/) (2025)

## Abstract (arXiv)

Kuaishou serves over 400 million daily active users, processing hundreds of millions of search queries daily against a repository of tens of billions of short videos. As the final decision layer, the reranking stage determines user experience by optimizing whole-page utility. While traditional score-and-sort methods fail to capture combinatorial dependencies, Generative Reranking offers a superior paradigm by directly modeling the permutation probability. However, deploying Generative Reranking in such a high-stakes environment faces a fundamental dual dilemma: 1) the structural trade-off where Autoregressive (AR) models offer superior Sequential modeling but suffer from prohibitive latency, versus Non-Autoregressive (NAR) models that enable efficiency but lack dependency capturing; 2) the optimization gap where Supervised Learning faces challenges in directly optimizing whole-page utility, while Reinforcement Learning (RL) struggles with instability in high-throughput data streams. To resolve this, we propose Dual-Rerank, a unified framework designed for industrial reranking that bridges the structural gap via Sequential Knowledge Distillation and addresses the optimization gap using List-wise Decoupled Reranking Optimization (LDRO) for stable online RL. Extensive A/B testing on production traffic demonstrates that Dual-Rerank achieves State-of-the-Art performance, significantly improving User satisfaction and Watch Time while drastically reducing inference latency compared to AR baselines.