In this study, we focus on the automatic evaluation of long and detailed image captions generated by multimodal Large Language Models (MLLMs). Most existing automatic evaluation metrics for image captioning are primarily designed for short captions and are not suitable for evaluating long captions. Moreover, recent LLM-as-a-Judge approaches suffer from slow inference due to their reliance on autoregressive inference and early fusion of visual information. To address these limitations, we propose VELA, an automatic evaluation metric for long captions developed within a novel LLM-Hybrid-as-a-Judge framework. Furthermore, we propose LongCap-Arena, a benchmark specifically designed for evaluating metrics for long captions. This benchmark comprises 7,805 images, the corresponding human-provided long reference captions and long candidate captions, and 32,246 human judgments from three distinct perspectives: Descriptiveness, Relevance, and Fluency. We demonstrated that VELA outperformed existing metrics and achieved superhuman performance on LongCap-Arena.

In recent years, large language models (LLMs) have demonstrated significant potential across various natural language processing (NLP) tasks. However, their performance in domain-specific applications and non-English languages remains less explored. This study introduces a novel Romanian-language dataset for multiple-choice biology questions, carefully curated to assess LLM comprehension and reasoning capabilities in scientific contexts. Containing approximately 14,000 questions, the dataset provides a comprehensive resource for evaluating and improving LLM performance in biology. We benchmark several popular LLMs, analyzing their accuracy, reasoning patterns, and ability to understand domain-specific terminology and linguistic nuances. Additionally, we perform comprehensive experiments to evaluate the impact of prompt engineering, fine-tuning, and other optimization techniques on model performance. Our findings highlight both the strengths and limitations of current LLMs in handling specialized knowledge tasks in low-resource languages, offering valuable insights for future research and development.

Recent advances in diffusion-based text-to-image (T2I) models have led to remarkable success in generating high-quality images from textual prompts. However, ensuring accurate alignment between the text and the generated image remains a significant challenge for state-of-the-art diffusion models. To address this, existing studies employ reinforcement learning with human feedback (RLHF) to align T2I outputs with human preferences. These methods, however, either rely directly on paired image preference data or require a learned reward function, both of which depend heavily on costly, high-quality human annotations and thus face scalability limitations. In this work, we introduce Text Preference Optimization (TPO), a framework that enables "free-lunch" alignment of T2I models, achieving alignment without the need for paired image preference data. TPO works by training the model to prefer matched prompts over mismatched prompts, which are constructed by perturbing original captions using a large language model. Our framework is general and compatible with existing preference-based algorithms. We extend both DPO and KTO to our setting, resulting in TDPO and TKTO. Quantitative and qualitative evaluations across multiple benchmarks show that our methods consistently outperform their original counterparts, delivering better human preference scores and improved text-to-image alignment. Our Open-source code is available at https://github.com/DSL-Lab/T2I-Free-Lunch-Alignment.

Large language models (LLMs) generate fluent text yet often default to safe, generic phrasing, raising doubts about their ability to handle creativity. We formalize this tendency as a Galton-style regression to the mean in language and evaluate it using a creativity stress test in advertising concepts. When ad ideas were simplified step by step, creative features such as metaphors, emotions, and visual cues disappeared early, while factual content remained, showing that models favor high-probability information. When asked to regenerate from simplified inputs, models produced longer outputs with lexical variety but failed to recover the depth and distinctiveness of the originals. We combined quantitative comparisons with qualitative analysis, which revealed that the regenerated texts often appeared novel but lacked true originality. Providing ad-specific cues such as metaphors, emotional hooks and visual markers improved alignment and stylistic balance, though outputs still relied on familiar tropes. Taken together, the findings show that without targeted guidance, LLMs drift towards mediocrity in creative tasks; structured signals can partially counter this tendency and point towards pathways for developing creativity-sensitive models.

Abstract--Despite progress in both traditional dexterous grasping pipelines and recent Vision-Language-Action (VLA) approaches, the grasp execution stage remains prone to pose inaccuracies, especially in long-horizon tasks, which undermines overall performance. T o address this "last-mile" challenge, we propose T acRefineNet, a tactile-only framework that achieves fine in-hand pose refinement of known objects in arbitrary target poses using multi-finger fingertip sensing. Our method iteratively adjusts the end-effector pose based on tactile feedback, aligning the object to the desired configuration. We design a multi-branch policy network that fuses tactile inputs from multiple fingers along with proprioception to predict precise control updates. T o train this policy, we combine large-scale simulated data from a physics-based tactile model in MuJoCo with real-world data collected from a physical system. Comparative experiments show that pretraining on simulated data and fine-tuning with a small amount of real data significantly improves performance over simulation-only training. T o our knowledge, this is the first method to enable arbitrary in-hand pose refinement via multi-finger tactile sensing alone. Project website is available at https://sites.google.com/view/tacrefinenet

Large Language Models (LLMs) have demonstrated outstanding performance in mathematical reasoning capabilities. However, we argue that current large-scale reasoning models primarily rely on scaling up training datasets with diverse mathematical problems and long thinking chains, which raises questions about whether LLMs genuinely acquire mathematical concepts and reasoning principles or merely remember the training data. In contrast, humans tend to break down complex problems into multiple fundamental atomic capabilities. Inspired by this, we propose a new paradigm for evaluating mathematical atomic capabilities. Our work categorizes atomic abilities into two dimensions: (1) field-specific abilities across four major mathematical fields, algebra, geometry, analysis, and topology, and (2) logical abilities at different levels, including conceptual understanding, forward multi-step reasoning with formal math language, and counterexample-driven backward reasoning. We propose corresponding training and evaluation datasets for each atomic capability unit, and conduct extensive experiments about how different atomic capabilities influence others, to explore the strategies to elicit the required specific atomic capability. Evaluation and experimental results on advanced models show many interesting discoveries and inspirations about the different performances of models on various atomic capabilities and the interactions between atomic capabilities. Our findings highlight the importance of decoupling mathematical intelligence into atomic components, providing new insights into model cognition and guiding the development of training strategies toward a more efficient, transferable, and cognitively grounded paradigm of "atomic thinking".

Abstract--Cellular wireless systems are witnessing the proliferation of frequency bands over a wide spectrum, particularly with the expansion of new bands in FR3. These bands must be supported in user equipment (UE) handsets with multiple antennas in a constrained form factor . Rapid variations in channel quality across the bands from motion and hand blockage, limited field-of-view of antennas, and hardware and power-constrained measurement sparsity pose significant challenges to reliable multi-band channel tracking. This paper formulates the problem of predicting achievable rates across multiple antenna arrays and bands with sparse historical measurements. We propose a transformer-based neural architecture that takes asynchronous rate histories as input and outputs per-array rate predictions. Evaluated on ray-traced simulations in a dense urban micro-cellular setting with FR1 and FR3 arrays, our method demonstrates superior performance over baseline predictors, enabling more informed band selection under realistic mobility and hardware constraints.

Time-series question answering (TSQA) tasks face significant challenges due to the lack of labeled data. Alternatively, with recent advancements in large-scale models, vision-language models (VLMs) have demonstrated the potential to analyze time-series signals in a zero-shot manner. In this paper, we propose a training approach that uses pseudo labels generated by a VLM. Although VLMs can produce incorrect labels, TSQA models can still be effectively trained based on the property that deep neural networks are inherently robust to such noisy labels. Our experimental results demonstrate that TSQA models are not only successfully trained with pseudo labels, but also surpass the performance of the VLM itself by leveraging a large amount of unlabeled data.

Many studies have shown various biases targeting different demographic groups in language models, amplifying discrimination and harming fairness. Recent parameter modification debiasing approaches significantly degrade core capabilities such as text coherence and task accuracy. And Prompt-based debiasing methods, only effective for predefined trigger words, fail to address deeply embedded stereotypical associations in model parameters. In this paper, we propose BiasUnlearn, a novel model debiasing framework which achieves targeted debiasing via dual-pathway unlearning mechanisms coordinating stereotype forgetting with anti-stereotype retention, while preventing bias polarity reversal through adversarial forget set and dynamic dataset swapping. We conducted extensive experiments with multiple language models across various evaluation benchmarks. The results show that BiasUnlearn outperforms existing methods in mitigating bias in language models while retaining language modeling capabilities. Further experiments reveal that debiasing weights are transferable across model variants, confirming that bias representations become entrenched during pre-training and persist through fine-tuning phases.

This paper presents an Artificial Intelligence (AI) integrated novel approach to Brain - Computer Interface (BCI) - based wheelchair development, utilizing a motor imagery r ight - l eft - h and m ovement mechanism for control. The system is designed to simulate wheelchair navigation based on motor imagery right and left - hand movements using electroencephalogram (EEG) data. A pre - filtered dataset, obtained from an open - source EEG repository, was seg mented into arrays of 19x200 to capture the onset of hand movements. Th e data was acquired at a sampling frequency of 200Hz. The system integrates a Tkinter - based interface for simulating wheelchair movements, offering users a functional and intuitive control system. We propose a BiLSTM - BiGRU model that shows a superior test accuracy of 92. 26 % as compared with v arious machine learning baseline models, including XGBoost, EEGNet, and a transformer - based model . The Bi - LSTM - BiGRU attention - based model achieved a mean accuracy of 90.13 % through cross - validation, showcasing the potential of attention mechanisms in BCI applications. Keywords: Brain Computer Interface (BCI), BiLSTM - BiGRU, Raspberry Pi, E lectroencephalogram (EEG), Hybrid Deep learning 1. Introduction Brain - Computer Interfaces (BCIs) are advanced systems that establish direct communication between the human brain and external devices . In recent years, BCIs have been widely investigated for their potential to assist individuals with mobility impairments, offering novel pathways for restoring autonomy. This paper proposes a BCI - based wheelchair control system driven by electroencephalogra phy (EEG) signals associated with motor imagery. The proposed framework incorporates a variety of machine learning models with tailored hyperparameter optimization techniques, culminating in the deployment of a BiLSTM - BiGRU hybrid deep learning model for effective EEG signal classification.