BLooP: Zero-Shot Abstractive Summarization using Large Language Models with Bigram Lookahead Promotion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Iyer, Varun, Caragea, Cornelia
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912962901442560
author Iyer, Varun
Caragea, Cornelia
author_facet Iyer, Varun
Caragea, Cornelia
contents Abstractive summarization requires models to generate summaries that convey information in the source document. While large language models can generate summaries without fine-tuning, they often miss key details and include extraneous information. We propose BLooP (Bigram Lookahead Promotion), a simple training-free decoding intervention that encourages large language models (LLMs) to generate tokens that form bigrams from the source document. BLooP operates through a hash table lookup at each decoding step, requiring no training, fine-tuning, or model modification. We demonstrate improvements in ROUGE and BARTScore for Llama-3.1-8B-Instruct, Mistral-Nemo-Instruct-2407, and Gemma-2-9b-it on CNN/DM, CCSum, Multi-News, and SciTLDR. Human evaluation shows that BLooP significantly improves faithfulness without reducing readability. We make the code available at https://github.com/varuniyer/BLooP
format Preprint
id arxiv_https___arxiv_org_abs_2603_11415
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle BLooP: Zero-Shot Abstractive Summarization using Large Language Models with Bigram Lookahead Promotion
Iyer, Varun
Caragea, Cornelia
Computation and Language
Abstractive summarization requires models to generate summaries that convey information in the source document. While large language models can generate summaries without fine-tuning, they often miss key details and include extraneous information. We propose BLooP (Bigram Lookahead Promotion), a simple training-free decoding intervention that encourages large language models (LLMs) to generate tokens that form bigrams from the source document. BLooP operates through a hash table lookup at each decoding step, requiring no training, fine-tuning, or model modification. We demonstrate improvements in ROUGE and BARTScore for Llama-3.1-8B-Instruct, Mistral-Nemo-Instruct-2407, and Gemma-2-9b-it on CNN/DM, CCSum, Multi-News, and SciTLDR. Human evaluation shows that BLooP significantly improves faithfulness without reducing readability. We make the code available at https://github.com/varuniyer/BLooP
title BLooP: Zero-Shot Abstractive Summarization using Large Language Models with Bigram Lookahead Promotion
topic Computation and Language
url https://arxiv.org/abs/2603.11415