A single deliberative quality score (AQuA) is calculated by combining predictions from adapter models trained on various aspects of deliberation and insights from both expert and non-expert annotations.
Rumor detection models struggle to detect unseen rumors due to overreliance on source post information and lack of consideration for contextual data.
LLMs struggle to match smaller models in zero-shot settings, prompting strategies impact accuracy significantly.
Large language models struggle to effectively distill text by removing forbidden variables while preserving other semantic content, posing challenges for computational social science investigations.
Social orientation tags improve dialogue outcome prediction and explainability.
Leveraging Large Language Models for simulating cultural evolution provides insights into human and machine-generated culture dynamics.
The author introduces the use of social orientation tags to predict and explain dialogue outcomes, demonstrating their effectiveness in improving task performance.