# GRASP: Deterministic argument ranking in interaction graphs

arXiv: https://arxiv.org/abs/2605.19141 · HTML: https://mlsanity.com/paper/2605.19141/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 33 of its 65 references that are in the map.
- First posted on arXiv 18 May 2026.

## Builds on (most cited first, 30 of 33)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models](https://mlsanity.com/paper/2512.02556/) (2025)
- [Adversarial NLI: A New Benchmark for Natural Language Understanding](https://mlsanity.com/paper/1910.14599/) (2019)
- [Justice or Prejudice? Quantifying Biases in LLM-as-a-Judge](https://mlsanity.com/paper/2410.02736/) (2024)
- [Probing Neural Network Comprehension of Natural Language Arguments](https://mlsanity.com/paper/1907.07355/) (2019)
- [Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges](https://mlsanity.com/paper/2406.12624/) (2024)
- [MiMo-V2-Flash Technical Report](https://mlsanity.com/paper/2601.02780/) (2026)
- [Large Language Models are Inconsistent and Biased Evaluators](https://mlsanity.com/paper/2405.01724/) (2024)
- [Aligning with Human Judgement: The Role of Pairwise Preference in Large Language Model Evaluators](https://mlsanity.com/paper/2403.16950/) (2024)
- [Systematic Biases in LLM Simulations of Debates](https://mlsanity.com/paper/2402.04049/) (2024)
- [Explaining Length Bias in LLM-Based Preference Evaluations](https://mlsanity.com/paper/2407.01085/) (2024)
- [Automatic Argument Quality Assessment -- New Datasets and Methods](https://mlsanity.com/paper/1909.01007/) (2019)
- [Exploring the Potential of Large Language Models in Computational Argumentation](https://mlsanity.com/paper/2311.09022/) (2023)
- [AMPERSAND: Argument Mining for PERSuAsive oNline Discussions](https://mlsanity.com/paper/2004.14677/) (2020)
- [Neither Valid nor Reliable? Investigating the Use of LLMs as Judges](https://mlsanity.com/paper/2508.18076/) (2025)
- [Exploring the Role of Prior Beliefs for Argument Persuasion](https://mlsanity.com/paper/1906.11301/) (2019)
- [Exploring the Role of Argument Structure in Online Debate Persuasion](https://mlsanity.com/paper/2010.03538/) (2020)
- [Are Large Language Models Reliable Argument Quality Annotators?](https://mlsanity.com/paper/2404.09696/) (2024)
- [A Corpus for Modeling User and Language Effects in Argumentation on Online Debating](https://mlsanity.com/paper/1906.11310/) (2019)
- [Can Language Models Recognize Convincing Arguments?](https://mlsanity.com/paper/2404.00750/) (2024)
- [A Comparative Study of Ranking-based Semantics for Abstract Argumentation](https://mlsanity.com/paper/1602.01059/) (2016)
- [Large Language Models in Argument Mining: A Survey](https://mlsanity.com/paper/2506.16383/) (2025)
- [Intrinsic Quality Assessment of Arguments](https://mlsanity.com/paper/2010.12473/) (2020)
- [Assessing the Sufficiency of Arguments through Conclusion Generation](https://mlsanity.com/paper/2110.13495/) (2021)
- [Equilibrium States in Numerical Argumentation Networks](https://mlsanity.com/paper/1408.6706/) (2014)
- [Can LLMs Judge Debates? Evaluating Non-Linear Reasoning via Argumentation Theory Semantics](https://mlsanity.com/paper/2509.15739/) (2025)
- [Inferring Attack Relations for Gradual Semantics](https://mlsanity.com/paper/2211.16118/) (2022)
- [Explain then Rank: Scale Calibration of Neural Rankers Using Natural Language Explanations from LLMs](https://mlsanity.com/paper/2402.12276/) (2024)

## Abstract (arXiv)

Large language models are increasingly deployed as automated judges to evaluate the strength of arguments. As this role expands, their legitimacy depends on consistency, transparency, and the ability to separate argumentative structure from rhetorical appeal. However, we show that holistic judging - a common LLM-as-a-Judge practice where a model provides a global verdict on a debate - suffers from substantial inter-model disagreement. We argue that this instability arises from collapsing a debate's complex interaction structure into a single opaque score. To address this, we propose GRASP (Gradual Ranking with Attacks and Support Propagation), a deterministic framework that aggregates stable local interaction judgments into a global ranking via a convergent attack--defense propagation operator. We show that local interaction judgments are more reproducible than holistic rankings in LLM-as-a-Judge evaluations, allowing GRASP to produce more consistent global rankings. We further show that GRASP scores do not correlate with human "convincingness" labels, highlighting a vital sociotechnical distinction: GRASP does not measure persuasion, factuality, or rhetorical appeal, but structural sufficiency - a defense-aware notion of argument robustness over the explicit interaction graph. Overall, GRASP offers a transparent and auditable alternative to holistic LLM judging.