Research map / Large language models / AI and society
PerSpectra: A Scalable and Configurable Pluralist Benchmark of Perspectives from Arguments
Shangrui Nie, Kian Omoomi, Lucie Flek, Zhixue Zhao, Charles Welch
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| compare human vs best model on opinion counting | LLMs · PerSpectra T1 Opinion Counting | accuracy: 36.2% -> 44.0% | Improved |
| compare human vs best model on opinion matching | LLMs · PerSpectra T2 Opinion Matching | accuracy: 85.4% -> 90.6% | Improved |
| compare human vs best model on polarity check | LLMs · PerSpectra T3 Polarity Check | accuracy: 76.4% -> 85.6% | Improved |
| scale Qwen3 from 8B to 32B | Qwen3 · PerSpectra T2 Opinion Matching | accuracy: 83.6% -> 71.6% | Hurt |
| scale Qwen3 from 8B to 32B | Qwen3 · PerSpectra T1 Opinion Counting | accuracy: 33.6% -> 28.6% | Hurt |
| scale Qwen3 from 8B to 32B | Qwen3 · PerSpectra T3 Polarity Check | accuracy: 64.0% -> 67.6% | Improved |
| compare stance accuracy to exact-match accuracy in opinion matching | LLMs · PerSpectra T2 Opinion Matching | accuracy: 83.6% -> 92.4% (Qwen3-8B) | Improved |
Closest papers by meaning
- Multi-LLM Debate: Framework, Principals, and Interventions (NeurIPS 2024 · similarity 0.88)
- Position: A Roadmap to Pluralistic Alignment (ICML 2024 · similarity 0.87)
- Debating with More Persuasive LLMs Leads to More Truthful Answers (ICML 2024 · similarity 0.87)
- Group Perspective Matters: Regulating Debate Relationships Can Mitigate Blind Conformity in Multi-Agent Debate (NeurIPS 2026 · similarity 0.86)
- Should we be going MAD? A Look at Multi-Agent Debate Strategies for LLMs (ICML 2024 · similarity 0.86)
- Representative Ranking for Deliberation in the Public Sphere (ICML 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
