---
title: "Diffusion-GR2: 扩散生成式推理重排序器"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.01170v1"
permalink: "https://feed.astralor.com/items/itm_373078b3c6a5e43c3c0f9e39"
published: "2026-07-01T17:02:20.000Z"
score: 59
channel: "ai"
tags: ["Diffusion-GR2","并行解码","扩散模型","推理重排序器","推荐系统"]
---

# Diffusion-GR2: 扩散生成式推理重排序器

## 速览

Diffusion-GR2将自回归推理重排序器转化为扩散模型，通过并行解码将推理吞吐量提升2.4-3.5倍，同时通过转换微调、在策略蒸馏和强化学习阶段弥合精度差距，在Amazon Beauty数据集上达到与自回归模型近似的推荐准确率。

## 关键点

- 提出Diffusion-GR2，将自回归推理重排序器转换为块扩散语言模型，实现并行解码加速。
- 通过转换微调解决结构差距，确保生成有效排序（无重复、遗漏或越界标识符）。
- 在策略蒸馏和强化学习阶段进一步弥合分布差距，恢复准确率。
- 在Amazon Beauty数据集上，推理吞吐量提升2.4-3.5倍，准确率接近自回归基线。

## 全文

Generative reasoning re-rankers achieve strong recommendation accuracy by emitting a chain-of-thought before re-ordering a candidate list, but they are slow at inference: an autoregressive (AR) decoder spends one sequential forward pass per reasoning token, and the reasoning trace far exceeds the ranking it produces. To reduce this cost, block-diffusion language models decode many positions in parallel over a few denoising steps and are substantially faster, yet naively converting an AR re-ranker into one opens two accuracy gaps: (1) a structural gap: answer positions are denoised in parallel and scored independently, so the decoder emits invalid rankings (duplicated, dropped, or out-of-set identifiers) that AR avoids through left-to-right masking; and (2) a distributional gap: fine-tuning the converted model on fixed teacher trajectories is off-policy relative to its own decoding at inference, leaving a residual accuracy gap. To close both gaps while keeping the speedup, we propose \textbf{Diffusion-GR2}, a recipe that converts our AR reasoning re-ranker (GR2) into a block-diffusion re-ranker. First, conversion fine-tuning (CFT) adapts the AR-initialized diffusion model to denoise the answer into a valid permutation on its own, without an external constrained decoder. Next, on-policy distillation (OPD) then supervises the model on its own decoded trajectories with dense per-token targets from the AR teacher. Finally, we apply a reinforcement-learning (RL) stage against a re-ranking reward on top of OPD's on-policy policy. Experiments on Amazon Beauty demonstrate that Diffusion-GR2 recovers to near-parity with the AR re-ranker, while block-parallel decoding raises decode throughput by $2.4$--$3.5\times$ at the model's reasoning output length. Ablations show that CFT recovers most of the conversion gap, and that on-policy distillation further closes it to the AR reference.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_373078b3c6a5e43c3c0f9e39