ACL Findings2021
LenAtten: An Effective Length Controlling Unit For Text Summarization
Zhongyi Yu, Zhenghao Wu, Hao Zheng, Zhe Xuanyuan, Jefferson Fong, Weifeng Su
Abstract
Fixed length summarization aims at generating summaries with a preset number of words or characters. Most recent researches incorporate length information with word embeddings as the input to the recurrent decoding unit, causing a compromise between length controllability and summary quality. In this work, we present an effective length controlling unit Length Attention (LenAtten) to break this trade-off. Experimental results show that LenAtten not only brings improvements in length controllability and ROGUE scores but also has great generalization ability. In the task of generating a summary with the target length, our model is 732 times better than the bestperforming length controllable summarizer in length controllability on the CNN/Daily Mail dataset. 1
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
5 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Length Controlled Generation for Black-box LLMs cites BERTScore: Evaluating Text Generation with BERT
- Length Controlled Generation for Black-box LLMs cites Length Control in Abstractive Summarization by Pretraining Information Selection
- Unsupervised Extractive Summarization with Learnable Length Control Strategies cites Extractive Summarization as Text Matching
- Unsupervised Extractive Summarization with Learnable Length Control Strategies cites Length Control in Abstractive Summarization by Pretraining Information Selection
- Hansel: Output Length Controlling Framework for Large Language Models cites Length Control in Abstractive Summarization by Pretraining Information Selection
Cited by top-tier papers5
- Length Control in Abstractive Summarization by Pretraining Information SelectionYizhu Liu, Qi Jia, Kenny Q. ZhuACL 2022 · 39 citations
- Unsupervised Extractive Summarization with Learnable Length Control StrategiesRenlong Jie, Xiaojun Meng, Xin Jiang, Qun LiuAAAI 2024 · 8 citations
- Hansel: Output Length Controlling Framework for Large Language ModelsSeoha Song, Junhyun Lee, Hyeonmok KoAAAI 2025 · 2 citations
- Length Controlled Generation for Black-box LLMsYuxuan Gu, Wenjie Wang, Xiaocheng Feng, Weihong Zhong et al.ACL 2025
- Fine-grained Video Dubbing Duration Alignment with Segment Supervised Preference OptimizationChaoqun Cui, Liangbin Huang, Shijing Wang, Zhe Tong et al.ACL 2025
Builds on2
Related papers
- A Decoding Algorithm for Length-Control Summarization Based on Directed Acyclic TransformersChenyang Huang, Hao Zhou, Cameron Jen, Kangjie Zheng et al.EMNLP Findings 2024
- A Character-Level Length-Control Algorithm for Non-Autoregressive Sentence SummarizationPuyuan Liu, Xiang Zhang, Lili MouNeurIPS 2022 · 21 citations
- Prompt-Based Length Controlled Generation with Multiple Control TypesRenlong Jie, Xiaojun Meng, Lifeng Shang, Xin Jiang et al.ACL Findings 2024 · 4 citations
- Syntactically Look-Ahead Attention Network for Sentence CompressionHidetaka Kamigaito, Manabu OkumuraAAAI 2020 · 22 citations
- DcLM: Output Length Control of Large Language Models via Dynamic Length MarkersZhe Chen, Jiaao Yu, Honglin LiACL Findings 2026
