Research map / Large language models / AI and society
Counterfactual LLM-based Framework for Measuring Rhetorical Style
Jingyi Qiu, Hong Chen, Zongyi Li
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 |
|---|---|---|---|
| use counterfactual BT rhetorical score as predictor instead of direct LLM rating | citation regression · ICLR 2017-2025 citations | regression coefficient: -26.11 -> 24.53 | Improved |
| use counterfactual BT rhetorical score instead of promotion lexicon score | media posts regression · ICLR posts | regression coefficient: 0.64 -> 3.19 | Improved |
| higher rhetorical score | paper attention · ICLR citations | regression coefficient: +24.53 citations per unit | Improved |
| higher rhetorical score | peer review rating · ICLR average reviewer score | Spearman correlation: rho = -0.015, p = 0.225 | No clear effect |
| increase number of personas from 8 to 15 in panel (ablation) | rhetorical scores · ICLR 8,485 abstracts | Spearman with full panel: 0.89 -> 0.958 | Improved |
| replace persona subset with a disjoint subset (ablation) | rhetorical scores · ICLR abstracts | Spearman correlation: mean 0.89 | No clear effect |
| year 2023 onward vs 2018-2022 | mean rhetorical score · ICLR submissions | rhetorical score | Hurt |
| higher estimated LLM usage in abstracts | rhetorical score · ICLR 2024-2025 | estimated LLM usage (%): 9.0% -> 20.9% | Improved |
Closest papers by meaning
- Understanding Annotator Perception: Modeling Psychological Inference from First- and Third-Person Annotations (Student Abstract) (AAAI 2025 · similarity 0.84)
- Certifying Counterfactual Bias in LLMs (ICLR 2025 · similarity 0.83)
- Uncovering Gaps in How Humans and LLMs Interpret Subjective Language (ICLR 2025 · similarity 0.83)
- Towards Unifying Evaluation of Counterfactual Explanations: Leveraging Large Language Models for Human-Centric Assessments (AAAI 2025 · similarity 0.83)
- Style Outweighs Substance: Failure Modes of LLM Judges in Alignment Benchmarking (ICLR 2025 · similarity 0.83)
- Metaphor Is Not All Attention Needs (NeurIPS 2026 · similarity 0.83)
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.
