This paper explores Chinese essay rhetoric recognition using Large Language Models (LLMs), LoRA, and in-context learning to assess linguistic and higher-order thinking skills. The proposed method achieved the best performance and won first prize in the CCL 2025 Chinese essay rhetoric recognition evaluation task.
SLAM (Structural Linguistic Activation Marking) is a novel white-box watermarking scheme for LLMs that embeds the mark into structural geometry rather than token frequencies. It achieves 100% detection accuracy with minimal quality loss, outperforming existing schemes.
This study proposes a structured framework to improve LLM reasoning when analyzing long documents, addressing issues like contextual bias and omission error. It combines parallel chunk-level processing with evidence-anchored consolidation to generate more robust and bias-resilient conceptual abstractions.
This study investigates the difficulty of interpreting peer review feedback, comparing the effectiveness of numerical scores versus text in predicting acceptance. The research reveals that score-based models are significantly more accurate (91%) than text-based models (81% even with LLMs), indicating textual information is considerably less reliable.
This paper proposes an evidence-based model to generate queries from query-free summarization datasets, addressing the challenge of finding suitable datasets for Query-Focused Summarization (QFS). Experimental results indicate that summaries generated using these evidence-based queries achieve competitive ROUGE scores, supporting their effectiveness for the QFS task.
This study employs Machine Learning, specifically the MARBERT model, to predict emoji use in Arabic tweets collected from X.com. The model achieved an overall accuracy of 0.75, indicating promising results while highlighting a need for further model improvement.
AdaGATE is a training-free evidence controller for multi-hop Retrieval-Augmented Generation (RAG) designed to address noisy or redundant retrieved evidence in limited contexts. It frames evidence selection as a token-constrained repair problem, combining entity-centric gap tracking and targeted micro-query generation to balance coverage, corroboration, and novelty.
This paper presents a comparative study of three explainability techniques (Integrated Gradients, Attention Rollout, and SHAP) on a fine-tuned DistilBERT model for sentiment classification. The study concludes that gradient-based attribution provides more stable and intuitive explanations, while attention-based methods are computationally efficient but less aligned with prediction-relevant features.
This paper introduces Hierarchical Policy Optimization (HPO) for Simultaneous Speech Translation (SST) using LLMs, addressing challenges like high computational cost and imperfect supervised fine-tuning data. HPO employs a hierarchical reward to balance translation quality and latency, demonstrating substantial improvements in COMET and MetricX scores.
This research evaluates cross-family speculative decoding for Polish LLMs on Apple Silicon, extending the MLX-LM framework with Universal Assisted Generation (UAG) for cross-tokenizer compatibility. Experiments show that context-aware token translation significantly improves acceptance rates for Bielik 11B on Polish language datasets.
This content discusses how local HVAC or plumbing business owners waste time manually deciphering technician voice notes full of jargon. It proposes using AI to automate this by training it to extract specific, structured data from unstructured speech, overcoming this business bottleneck.
CFMS introduces the first fine-grained Chinese multimodal sarcasm detection benchmark, comprising 2,796 image-text pairs with triple-level annotations. This dataset aims to improve AI's fine-grained semantic understanding and metaphoric reasoning, addressing limitations in existing benchmarks.
Transformers struggle with high computational costs and memory consumption for long sequences, while alternatives lose long-tail dependencies. Absorber LLM proposes a self-supervised causal synchronization to absorb historical contexts into parameters, ensuring a contextless model matches the original full-context one on future generations.
LiFT is a new instruction fine-tuning framework designed to improve in-context learning for large language models on longitudinal NLP tasks, which require reasoning over temporally ordered text. It uses a curriculum that progressively increases temporal difficulty, incorporating few-shot structure and temporal conditioning, consistently outperforming base models across various datasets and parameter sizes.
This paper introduces Derivation Prompting, a novel prompting technique for the Retrieval-Augmented Generation (RAG) framework. The method aims to reduce hallucinations and erroneous reasoning in Large Language Models (LLMs) by systematically applying predefined rules to derive conclusions. A case study demonstrated a significant reduction in unacceptable answers compared to traditional RAG methods.
This paper details participation in SemEval-2026 Task 13, focusing on lightweight detection of LLM-generated code using stylometric signals. The approach employs ratio-based features, parsing engines, and language classifiers, proving computationally efficient with near-instant inference time.
This paper evaluates the ability of large language models (LLMs) to adapt their responses to the certainty of retrieved information, revealing systematic limitations. It proposes an interaction strategy combining prior reminders, certainty recalibration, and context simplification to enhance LLM reliability. This approach reduces obedience errors by 25% without modifying model weights.
The research explores the capabilities of lightweight Large Language Models (LLMs) in Criminal Court View Generation (CVG) and their impact on charge prediction within Legal AI. It systematically investigates architectural differences, model size, and comparison with Deep Neural Networks, introducing the CVGEvalKit framework for evaluation.
MultiSoc-4D is a new Bengali social media dataset benchmark designed to diagnose LLM behavior in closed-set annotation. The research identifies "instruction-induced label collapse," a phenomenon where LLMs systematically prefer fallback labels, leading to under-detection of minority categories.
This research introduces a scalable computational approach to measure manner and result verbs, a crucial distinction for developmental language studies. It leverages large language models for sentence annotations and trains a RoBERTa-based classifier, demonstrating promising performance on evaluation datasets.