# Rethinking Model Selection and Decoding for Keyphrase Generation with Pre-trained Sequence-to-Sequence Models

arXiv: https://arxiv.org/abs/2310.06374 · HTML: https://mlsanity.com/paper/2310.06374/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning.
- Builds on 41 of its 65 references that are in the map.
- First posted on arXiv 10 Oct 2023.

## Placed in

- LLM reasoning / [Test-time compute and efficient thinking](https://mlsanity.com/lr/test-time-compute-and-efficient-thinking/)

## Builds on (most cited first, 30 of 41)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://mlsanity.com/paper/1910.13461/) (2019)
- [Scaling Instruction-Finetuned Language Models](https://mlsanity.com/paper/2210.11416/) (2022)
- [The Curious Case of Neural Text Degeneration](https://mlsanity.com/paper/1904.09751/) (2019)
- [BioBERT: a pre-trained biomedical language representation model for biomedical text mining](https://mlsanity.com/paper/1901.08746/) (2019)
- [Don't Stop Pretraining: Adapt Language Models to Domains and Tasks](https://mlsanity.com/paper/2004.10964/) (2020)
- [SciBERT: A Pretrained Language Model for Scientific Text](https://mlsanity.com/paper/1903.10676/) (2019)
- [Get To The Point: Summarization with Pointer-Generator Networks](https://mlsanity.com/paper/1704.04368/) (2017)
- [Hierarchical Neural Story Generation](https://mlsanity.com/paper/1805.04833/) (2018)
- [What Does BERT Look At? An Analysis of BERT's Attention](https://mlsanity.com/paper/1906.04341/) (2019)
- [Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks](https://mlsanity.com/paper/2204.07705/) (2022)
- [Cross-Task Generalization via Natural Language Crowdsourcing Instructions](https://mlsanity.com/paper/2104.08773/) (2021)
- [Quantifying the Carbon Emissions of Machine Learning](https://mlsanity.com/paper/1910.09700/) (2019)
- [Analyzing the Structure of Attention in a Transformer Language Model](https://mlsanity.com/paper/1906.04284/) (2019)
- [Deep Keyphrase Generation](https://mlsanity.com/paper/1704.06879/) (2017)
- [One Size Does Not Fit All: Generating and Evaluating Variable Number of Keyphrases](https://mlsanity.com/paper/1810.05241/) (2018)
- [Unsupervised Keyphrase Extraction with Multipartite Graphs](https://mlsanity.com/paper/1803.08721/) (2018)
- [Keyphrase Generation with Correlation Constraints](https://mlsanity.com/paper/1808.07185/) (2018)
- [Neural Keyphrase Generation via Reinforcement Learning with Adaptive Rewards](https://mlsanity.com/paper/1906.04106/) (2019)
- [One2Set: Generating Diverse Keyphrases as a Set](https://mlsanity.com/paper/2105.11134/) (2021)
- [Exclusive Hierarchical Decoding for Deep Keyphrase Generation](https://mlsanity.com/paper/2004.08511/) (2020)
- [Semi-Supervised Learning for Neural Keyphrase Generation](https://mlsanity.com/paper/1808.06773/) (2018)
- [Title-Guided Encoding for Keyphrase Generation](https://mlsanity.com/paper/1808.08575/) (2018)
- [Learning Rich Representation of Keyphrases from Text](https://mlsanity.com/paper/2112.08547/) (2021)
- [An Integrated Approach for Keyphrase Generation via Exploring the Power of Retrieval and Extraction](https://mlsanity.com/paper/1904.03454/) (2019)
- [An Empirical Study on Neural Keyphrase Generation](https://mlsanity.com/paper/2009.10229/) (2020)
- [Select, Extract and Generate: Neural Keyphrase Generation with Layer-wise Coverage Attention](https://mlsanity.com/paper/2008.01739/) (2020)
- [Keyphrase Generation for Scientific Document Retrieval](https://mlsanity.com/paper/2106.14726/) (2021)

## Built on by (most cited first, 8 of 8)

- [Unsupervised Domain Adaptation for Keyphrase Generation using Citation Contexts](https://mlsanity.com/paper/2409.13266/) (2024)
- [On Leveraging Encoder-only Pre-trained Language Models for Effective Keyphrase Generation](https://mlsanity.com/paper/2402.14052/) (2024)
- [Pre-Trained Language Models for Keyphrase Prediction: A Review](https://mlsanity.com/paper/2409.01087/) (2024)
- [One2set + Large Language Model: Best Partners for Keyphrase Generation](https://mlsanity.com/paper/2410.03421/) (2024)
- [Self-Compositional Data Augmentation for Scientific Keyphrase Generation](https://mlsanity.com/paper/2411.03039/) (2024)
- [MetaKP: On-Demand Keyphrase Generation](https://mlsanity.com/paper/2407.00191/) (2024)
- [Zero-Shot Keyphrase Generation: Investigating Specialized Instructions and Multi-Sample Aggregation on Large Language Models](https://mlsanity.com/paper/2503.00597/) (2025)
- [An Analysis of Datasets, Metrics and Models in Keyphrase Generation](https://mlsanity.com/paper/2506.10346/) (2025)

## Abstract (arXiv)

Keyphrase Generation (KPG) is a longstanding task in NLP with widespread applications. The advent of sequence-to-sequence (seq2seq) pre-trained language models (PLMs) has ushered in a transformative era for KPG, yielding promising performance improvements. However, many design decisions remain unexplored and are often made arbitrarily. This paper undertakes a systematic analysis of the influence of model selection and decoding strategies on PLM-based KPG. We begin by elucidating why seq2seq PLMs are apt for KPG, anchored by an attention-driven hypothesis. We then establish that conventional wisdom for selecting seq2seq PLMs lacks depth: (1) merely increasing model size or performing task-specific adaptation is not parameter-efficient; (2) although combining in-domain pre-training with task adaptation benefits KPG, it does partially hinder generalization. Regarding decoding, we demonstrate that while greedy search achieves strong F1 scores, it lags in recall compared with sampling-based methods. Based on these insights, we propose DeSel, a likelihood-based decode-select algorithm for seq2seq PLMs. DeSel improves greedy search by an average of 4.7% semantic F1 across five datasets. Our collective findings pave the way for deeper future investigations into PLM-based KPG.