Research map / Large language models / AI and society
Characterizing AI Manipulation Risks in Brazilian YouTube Climate Discourse
Wenchao Dong, Marcelo Sartori Locatelli, Virgilio Almeida, Meeyoung Cha
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 |
|---|---|---|---|
| fine-tune comment generator on top-liked comments instead of random comments (ablation) | Llama-3-8B comment generator · Comment generation, K=1 | estimated likes: 2.20 -> 7.25 | Improved |
| generate comments with intention ToM profile instead of random baseline | Llama-3-8B comment generator · Comment generation, K=1 | estimated likes: 2.20 -> 2.26 | No clear effect |
| fine-tune on extreme denial narratives rather than general denial comments | Denial comment generator · Comment generation, K=1 | estimated likes: 1.91 -> 2.37 | Improved |
| add video context to LLM judge for popularity prediction (ablation) | GPT-4.1 · Relative likability prediction | accuracy: Rand. 0.75 -> 0.76; Vid. 0.77 -> 0.81 | Mixed |
| add emotional ToM narratives to popularity prediction (ablation) | likability predictors · Relative likability prediction | accuracy: average improvement of 4.69% | Improved |
| use moral appeal strategy in climate videos | video like ratio regression · Brazilian climate YouTube videos | like ratio: average increase of 2.1% in video likes | Improved |
| use logical appeal, authority or common ground strategies | video like ratio regression · Brazilian climate YouTube videos | like ratio | Hurt |
| use emotional appeal strategy | video like ratio regression · Brazilian climate YouTube videos | like ratio | Improved |
Closest papers by meaning
- Assessing Large Language Models on Climate Information (ICML 2024 · similarity 0.85)
- A Multimodal Benchmark for Framing of Oil & Gas Advertising and Potential Greenwashing Detection (NeurIPS 2025 · similarity 0.85)
- Measuring And Improving Persuasiveness Of Large Language Models (ICLR 2025 · similarity 0.84)
- DEBATE: A Large-Scale Benchmark for Evaluating Opinion Dynamics in Role-Playing LLM Agents (NeurIPS 2026 · similarity 0.81)
- Examining Alignment of Large Language Models through Representative Heuristics: the case of political stereotypes (ICLR 2025 · similarity 0.81)
- InvestESG: A multi-agent reinforcement learning benchmark for studying climate investment as a social dilemma (ICLR 2025 · similarity 0.81)
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.
