---
title: "Training-Free Character-Length Control in Summarization with Diffusion Language Models"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.06371v1"
permalink: "https://feed.astralor.com/items/itm_fe1d54f924ef312204622a91"
published: "2026-11-04T11:32:20.000Z"
score: 56
channel: "tech"
tags: ["推理效率","Training-Free Character-Length Control","Summarization","Diffusion Language Models","Autoregressive","LMs"]
---

# Training-Free Character-Length Control in Summarization with Diffusion Language Models

## 速览

Summarization with a fixed character length is important for tasks such as headline and article preview generation, where text must fit an exact space. Autoregressive language models (LMs) struggle with precise characte…

## 关键点

- Summarization with a fixed character length is important for tasks such as headline and article preview generation, where text must fit an exact space.
- Autoregressive language models (LMs) struggle with precise character-length control: training-based methods require task-specific fine-tuning, while zero-shot prompting strategies achieve only approximate compliance.
- We propose DELTA (Diffusion Expected Length Targeting Algorithm), a training-free inference-time method for masked diffusion LMs that estimates expected character length at each denoising step and dynamically inserts or…

## 全文

Summarization with a fixed character length is important for tasks such as headline and article preview generation, where text must fit an exact space. Autoregressive language models (LMs) struggle with precise character-length control: training-based methods require task-specific fine-tuning, while zero-shot prompting strategies achieve only approximate compliance. We propose DELTA (Diffusion Expected Length Targeting Algorithm), a training-free inference-time method for masked diffusion LMs that estimates expected character length at each denoising step and dynamically inserts or removes masked positions to steer toward a target length. DELTA requires no fine-tuning and is designed for off-the-shelf masked diffusion LMs. Across three summarization benchmarks, it reduces character-length errors by an order of magnitude compared to baselines while maintaining competitive summary quality in both automatic and human evaluation.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_fe1d54f924ef312204622a91