The detection of climate misinformation narratives remains challenging due to label imbalance, hierarchical taxonomies, and the multi-label nature of real-world claims. Developing models that can reliably assign fine-grained narrative categories is therefore essential for scalable analysis of climate disinformation. We present our approach to multi-label climate misinformation narrative classification for ClimateCheck@NSLP 2026 Task 2. The task requires assigning one or more narrative categories, defined by the hierarchical CARDS taxonomy, to climate-related claims. We investigate both encoder-based transformers and decoder-only large language models (LLMs), comparing fine-tuning BERT-based models with prompt-based and retrieval-augmented instruction tuning strategies with Qwen3 model. To address data scarcity and label imbalance, we explore targeted augmentation using external CARDS-based resources as well as semantic similarity filtering. Our experiments show that augmentation improves encoder-based models, with ModernBERT achieving competitive performance at low computational cost. However, the strongest results are obtained using retrieval-augmented instruction tuning with Qwen3, which narrows the candidate narrative space prior to prediction. This approach achieves a Macro-F1 score of 59.72% on the official test set, securing second place on the leaderboard. These findings demonstrate the effectiveness of retrieval-guided LLM adaptation for structured multi-label narrative classification while highlighting the continued relevance of efficient encoder-based models.

This paper presents an efficiency-aware pipeline for automated fact-checking of real-world image–text claims that treats multimodality as a controllable design variable rather than a property that must be uniformly propagated through every stage of the system. The approach decomposes claims into verification questions, assigns each to text- or image-related types, and applies modality-aware retrieval strategies, while ultimately relying on text-only evidence for verdict prediction and justification generation. Evaluated on the AVerImaTeC dataset within the FEVER-9 shared task, the system achieves competitive question, evidence, verdict, and justification scores and ranks fourth overall, outperforming the official baseline on evidence recall, verdict accuracy, and justification quality despite not using visual evidence during retrieval. These results demonstrate that strong performance on multimodal fact-checking can be achieved by selectively controlling where visual information influences retrieval and reasoning, rather than performing full multimodal fusion at every stage of the pipeline.

Temporal language does more than place events on a timeline. In news discourse, references to the past, present, and future can function as rhetorical devices that shape interpretation and persuasion. Here, we study temporal framing, defined as the persuasive use of time-related language to structure meaning rather than to report chronology. We propose a taxonomy of eight temporal frames grounded in prior work on temporality and framing, and we realize it through expert annotation of a multilingual news corpus. The resulting dataset includes 458 English and German news articles, with over 2K temporally framed sentences and approximately 3K temporal framing annotations identified from a corpus of more than 20K sentences. We analyze frame prevalence, co-occurrence patterns, and lexical cues, and evaluate temporal framing detection using supervised fine-tuning and zero-shot classification. Our experiments show that temporal framing is learnable at the sentence level, with supervised models substantially outperforming zero-shot approaches. We publicly release the corpus to support future research on temporal framing: https://mbzuai-nlp.github.io/temporal-framing/.