Tag
This paper conducts a multi-method computational analysis of Telegram discourse on the Israel-Palestine conflict, revealing that pro-Israel channels use a neutral, report-style tone while pro-Palestine channels exhibit more negative sentiment and framing.
This paper proposes a rationale-guided knowledge distillation framework for cross-lingual stance detection, using chain-of-thought prompting from large language models to train a compact student model with dual-path distillation and contrastive learning.
PAST-TIDE is a stance detection system for the StanceNakba Shared Task, using statement tuning with cloze-style masked language modeling, prototypical contrastive learning, and topic-conditional layer normalization for cross-topic Arabic stance detection, achieving macro-F1 scores of 0.75 and 0.74 on subtasks A and B.
This paper proposes KIRP, a zero-shot stance detection framework for tweets that integrates external knowledge with entity reorganization and reflective chain-of-thought reasoning, achieving state-of-the-art performance on multiple datasets including a newly constructed Japanese tweet dataset.
This paper empirically measures how ten linguistic features in fine-tuning data shift Llama-3.2-1B's reasoning on animal welfare, finding that assertive and moral language strengthens pro-animal-welfare stances while hedged and descriptive language dilutes them.
This paper presents a method using LLMs for stance detection in scientific discourse, specifically identifying realism vs. instrumentalism in Bayesian cognitive science articles. The approach combines theory-driven coding, expert annotations, and prompt optimization to achieve high reliability.
Introduces SICI, a seven-dimensional diagnostic measure to assess semantic-pragmatic complexity in stance detection for LLMs, revealing regime shifts in error patterns across models and prompting strategies.
Presents BioStance, a context-aware dataset of 39,600 annotated Reddit post-comment pairs for stance detection in bioethical controversies, covering six targets across three dimensions of bioethical debate.
Introduces a multi-agent reasoning framework for stance detection that uses a Manager–Worker architecture with adaptive worker allocation. The framework achieves strong results on implicit stance cases, outperforming baselines on COVID-19 and SemEval-2016 datasets.
This paper proposes an Interpretive Audit Pipeline that leverages multi-model disagreement to detect interpretive complexity in LLM-based public comment analysis, arguing that disagreement-based evaluation is a necessary complement to standard accuracy metrics.
Researchers from the University of British Columbia propose an unsupervised graph-based system for organizing arguments from online debates by constructing interaction graphs and applying community detection to reveal diverse viewpoint distributions. The approach requires no training data and aims to help users navigate complex argumentative landscapes and combat filter bubbles.