# Do LLM Debates Repeat Arguments Differently Across Languages?

arXiv: https://arxiv.org/abs/2607.23442 · HTML: https://mlsanity.com/paper/2607.23442/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 35 of its 62 references that are in the map.
- First posted on arXiv 26 Jul 2026.

## Builds on (most cited first, 30 of 35)

- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [The Curious Case of Neural Text Degeneration](https://mlsanity.com/paper/1904.09751/) (2019)
- [Improving Factuality and Reasoning in Language Models through Multiagent Debate](https://mlsanity.com/paper/2305.14325/) (2023)
- [A Diversity-Promoting Objective Function for Neural Conversation Models](https://mlsanity.com/paper/1510.03055/) (2015)
- [M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation](https://mlsanity.com/paper/2402.03216/) (2024)
- [Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate](https://mlsanity.com/paper/2305.19118/) (2023)
- [Language-agnostic BERT Sentence Embedding](https://mlsanity.com/paper/2007.01852/) (2020)
- [Language Models are Multilingual Chain-of-Thought Reasoners](https://mlsanity.com/paper/2210.03057/) (2022)
- [Neural Text Generation with Unlikelihood Training](https://mlsanity.com/paper/1908.04319/) (2019)
- [AI safety via debate](https://mlsanity.com/paper/1805.00899/) (2018)
- [Multilingual E5 Text Embeddings: A Technical Report](https://mlsanity.com/paper/2402.05672/) (2024)
- [Diverse Beam Search: Decoding Diverse Solutions from Neural Sequence Models](https://mlsanity.com/paper/1610.02424/) (2016)
- [Exploring Collaboration Mechanisms for LLM Agents: A Social Psychology View](https://mlsanity.com/paper/2310.02124/) (2023)
- [Not All Languages Are Created Equal in LLMs: Improving Multilingual Capability by Cross-Lingual-Thought Prompting](https://mlsanity.com/paper/2305.07004/) (2023)
- [Having Beer after Prayer? Measuring Cultural Bias in Large Language Models](https://mlsanity.com/paper/2305.14456/) (2023)
- [Probing Pre-Trained Language Models for Cross-Cultural Differences in Values](https://mlsanity.com/paper/2203.13722/) (2022)
- [Learning to Break the Loop: Analyzing and Mitigating Repetitions for Neural Text Generation](https://mlsanity.com/paper/2206.02369/) (2022)
- [Exchange-of-Thought: Enhancing Large Language Model Capabilities through Cross-Model Communication](https://mlsanity.com/paper/2312.01823/) (2023)
- [Can multiple-choice questions really be useful in detecting the abilities of LLMs?](https://mlsanity.com/paper/2403.17752/) (2024)
- [Talk Isn't Always Cheap: Understanding Failure Modes in Multi-Agent Debate](https://mlsanity.com/paper/2509.05396/) (2025)
- [Can LLM Agents Really Debate? A Controlled Study of Multi-Agent Debate in Logical Reasoning](https://mlsanity.com/paper/2511.07784/) (2025)
- [Break the Checkbox: Challenging Closed-Style Evaluations of Cultural Alignment in LLMs](https://mlsanity.com/paper/2502.08045/) (2025)
- [Measuring the Similarity of Sentential Arguments in Dialog](https://mlsanity.com/paper/1709.01887/) (2017)
- [Stay Focused: Problem Drift in Multi-Agent Debate](https://mlsanity.com/paper/2502.19559/) (2025)
- [Debatrix: Multi-dimensional Debate Judge with Iterative Chronological Analysis Based on LLM](https://mlsanity.com/paper/2403.08010/) (2024)
- [Multiple LLM Agents Debate for Equitable Cultural Alignment](https://mlsanity.com/paper/2505.24671/) (2025)
- [Could Thinking Multilingually Empower LLM Reasoning?](https://mlsanity.com/paper/2504.11833/) (2025)
- [Debate, Deliberate, Decide (D3): A Cost-Aware Adversarial Framework for Reliable and Interpretable LLM Evaluation](https://mlsanity.com/paper/2410.04663/) (2024)
- [Do Repetitions Matter? Strengthening Reliability in LLM Evaluations](https://mlsanity.com/paper/2509.24086/) (2025)

## Abstract (arXiv)

LLM debate is usually evaluated by final answers, yet transcripts reveal whether later turns develop new arguments or return to earlier claims in new wording. We study this process with textitprior-argument similarity, which compares extracted argument units with earlier units in the same debate. In controlled eight-turn debates over 71 motions, six languages, and four model agents, Chinese is the only tested language with a consistently positive gap relative to English across three multilingual embedding models. The gap persists across agents, turn positions, regression adjustment, metric variants, extraction-length controls, a second-extractor subset, and cross-encoder tail rescoring. Manual calibration shows weak item-level alignment but a high-similarity tail enriched for substantive repetition. A diversity-aware prompt lowers textitprior-argument similarity across languages, yet does not significantly narrow the Chinese--English gap. Multilingual debate evaluation should therefore measure argumentative development over time and report both average and gap terms.