Zhicong Li, Lingjie Jiang, Yulan Hu, Xingchen Zeng, Yixia Li, Xiangwen Zhang, Guanhua Chen, Zheng Pan, Xin Li, Yong Liu. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Yixia Li, Yaqing Shi, Zhiwen Ruan, Dongdong Zhang, Lingjie Jiang, Shaohan Huang, Yun Chen, Guanhua Chen, Furu Wei. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Di Zhang, Xun Wu, Shaohan Huang, Lingjie Jiang, Yaru Hao, Li Dong, Zewen Chi, Zhifang Sui, Furu Wei. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Image aesthetics is a crucial metric in the field of image generation.However, textual aesthetics has not been sufficiently explored.With the widespread application of large language models (LLMs), previous work has primarily focused on the correctness of content and the helpfulness of responses.Nonetheless, providing responses with textual aesthetics is also an important factor for LLMs, which can offer a cleaner layout and ensure greater consistency and coherence in content.In this work, we introduce a pipeline for aesthetics polishing and help construct a textual aesthetics dataset named TEXAES.We propose a textual aesthetics-powered fine-tuning method based on direct preference optimization, termed TAPO, which leverages textual aesthetics without compromising content correctness.Additionally, we develop two evaluation methods for textual aesthetics based on text and image analysis, respectively.Our experiments demonstrate that using textual aesthetics data and employing the TAPO fine-tuning method not only improves aesthetic scores but also enhances performance on general evaluation datasets such as AlpacalEval and Arena-hard.Our code and data are available at https://github.com/ JackLingjie/

Recent Large Reasoning Models (LRMs) have shown substantially improved reasoning capabilities over traditional Large Language Models (LLMs) by incorporating extended thinking processes prior to producing final responses. However, excessively lengthy thinking introduces substantial overhead in terms of token consumption and latency, which is unnecessary for simple queries. In this work, we introduce Large Hybrid-Reasoning Models (LHRMs), the first kind of model capable of adaptively determining whether to perform reasoning based on the contextual information of user queries. To achieve this, we propose a two-stage training pipeline comprising Hybrid Fine-Tuning (HFT) as a cold start, followed by online reinforcement learning with the proposed Hybrid Group Policy Optimization (HGPO) to implicitly learn to select the appropriate reasoning mode. Furthermore, we introduce a metric called Hybrid Accuracy to quantitatively assess the model’s capability for hybrid reasoning. Extensive experimental results show that LHRMs can adaptively perform hybrid reasoning on queries of varying difficulty and type. It outperforms existing LRMs and LLMs in reasoning and general capabilities while significantly improving efficiency. Together, our work advocates for a reconsideration of the appropriate use of extended reasoning processes and provides a solid starting point for building hybrid reasoning systems.

In the past decade, object detection technology has developed rapidly. However, in the field of unmanned aerial vehicle (UAV) image object detection, challenges such as complex environments, numerous and dense small objects, and weak features make object detection from the UAV perspective a highly challenging task. To address these issues, this letter proposes a U-Shaped feature extraction and fusion network (U-ShapeNet). Specifically: first, to enhance the network’s feature extraction capability and improve the perception of small objects, we design a novel U-Shaped feature extraction network (U-SFEN) and introduce a tiny object detection head. Second, a large kernel feature selection module (LKFSM) is constructed to strengthen the network’s contextual information learning ability and effectively distinguish small objects from complex background noise. Third, a same-scale feature enhancement module (SFEM) is proposed to mitigate information decay by reusing same-scale feature maps. Experiments on the VisDrone2019 and HazyDet datasets demonstrate that U-ShapeNet outperforms current mainstream object detectors, achieving state-of-the-art performance.