mlsanityData to 7 Oct 2026

Arg-LLaDA: Argument Summarization via Large Language Diffusion Models and Sufficiency-Aware Refinement

arXiv 2507.19081 ↗First posted 25 Jul 2025cs.CLcited by 0

On the map: Diffusion models → Discrete and text diffusion

Abstract (arXiv)

Argument summarization aims to generate concise, structured representations of complex, multi-perspective debates. While recent work has advanced the identification and clustering of argumentative components, the generation stage remains underexplored. Existing approaches typically rely on single-pass generation, offering limited support for factual correction or structural refinement. To address this gap, we introduce Arg-LLaDA, a novel large language diffusion framework that iteratively improves summaries via sufficiency-guided remasking and regeneration. Our method combines a flexible masking controller with a sufficiency-checking module to identify and revise unsupported, redundant, or incomplete spans, yielding more faithful, concise, and coherent outputs. Empirical results on two benchmark datasets demonstrate that Arg-LLaDA surpasses state-of-the-art baselines in 7 out of 10 automatic evaluation metrics. In addition, human evaluations reveal substantial improvements across core dimensions, coverage, faithfulness, and conciseness, validating the effectiveness of our iterative, sufficiency-aware generation strategy.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Discrete and text diffusion, one of 14 directions in Diffusion models. Builds on 34 of its 58 references that are in the map. First posted on arXiv 25 Jul 2025.

Builds on

34 of 58 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024Simplified and Generalized Masked Diffusion for Discrete Data6 Jun
  2. 2024Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Data6 Jun

Other references · 32

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  4. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
  5. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,928
  6. 2020BLEURT: Learning Robust Metrics for Text Generationcited by 1,196
  7. 2025LLaDA-V: Large Language Diffusion Models with Visual Instruction Tuningcited by 145
  8. 2023Factually Consistent Summarization via Reinforcement Learning with Textual Entailment Feedbackcited by 71
  9. 2020From Arguments to Key Points: Towards Automatic Argument Summarizationcited by 60
  10. 2019Classification and Clustering of Arguments with Contextualized Word Embeddingscited by 57
  11. 2016Neural Network-Based Abstract Generation for Opinions and Argumentscited by 46
  12. 2020Aspect-Controlled Neural Argument Generationcited by 43
20 more
  1. 2023ArguGPT: evaluating, understanding and identifying argumentative essays generated by GPT modelscited by 41
  2. 2022Mining Legal Arguments in Court Decisionscited by 35
  3. 2023SEEDS: Exponential SDE Solvers for Fast High-Quality Sampling from Diffusion Modelscited by 35
  4. 2020Quantitative Argument Summarization and Beyond: Cross-Domain Key Point Analysiscited by 31
  5. 2022Towards Improving Faithfulness in Abstractive Summarizationcited by 27
  6. 2022IAM: A Comprehensive and Large-Scale Dataset for Integrated Argument Mining Taskscited by 23
  7. 2025MIRA: Medical Time Series Foundation Model for Real-World Health Datacited by 21
  8. 2021Every Bite Is an Experience: Key Point Analysis of Business Reviewscited by 19
  9. 2025BRIDGE: Bootstrapping Text to Control Time-Series Generation via Multi-Agent Iterative Optimization and Diffusion Modelingcited by 18
  10. 2021Key Point Analysis via Contrastive Learning and Extractive Argument Summarizationcited by 16
  11. 2023Augmenting LLMs with Knowledge: A survey on hallucination preventioncited by 16
  12. 2025Large Language Models in Argument Mining: A Surveycited by 14
  13. 2024Which Side Are You On? A Multi-task Dataset for End-to-End Argument Summarisation and Evaluationcited by 13
  14. 2021Team Enigma at ArgMining-EMNLP 2021: Leveraging Pre-trained Language Models for Key Point Matchingcited by 11
  15. 2024An Empirical Analysis of Diversity in Argument Summarizationcited by 9
  16. 2024Prompted Aspect Key Point Analysis for Quantitative Review Summarizationcited by 6
  17. 2024Enhancing Argument Summarization: Prioritizing Exhaustiveness in Key Point Generation and Introducing an Automatic Coverage Evaluation Metriccited by 5
  18. 2024Exploring Key Point Analysis with Pairwise Generation and Graph Partitioningcited by 4
  19. 2025Argument Summarization and its Evaluation in the Era of Large Language Modelscited by 4
  20. 2017Summarizing Dialogic Arguments from Social Mediacited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026ArgBench: Benchmarking LLMs on Computational Argumentation Tasks
  2. 2023Argue with Me Tersely: Towards Sentence-Level Counter-Argument Generation
  3. 2024ConQRet: Benchmarking Fine-Grained Evaluation of Retrieval Augmented Argumentation with LLM Judges
  4. 2025Limited Generalizability in Argument Mining: State-Of-The-Art Models Learn Datasets, Not Arguments
  5. 2025DebateBench: A Challenging Long Context Reasoning Benchmark For Large Language Models
  6. 2025Can LLMs Judge Debates? Evaluating Non-Linear Reasoning via Argumentation Theory Semantics
  7. 2026From Argument Components to Graphs: A Multi-Agent Debate with Confidence Gating for Argument Relations
  8. 2026Argument Quality Assessment with Large Language Models: A Pairwise Bradley-Terry Approach