Shuzhang Cai

Zhiqiang Wang, Pengbin Feng, Yanbin Lin et al.

We propose Fuzzy Group Relative Policy Reward (FGRPR), a novel framework that integrates Group Relative Policy Optimization (GRPO) with a fuzzy reward function to enhance learning efficiency. Unlike the conventional binary 0/1 accuracy reward, our fuzzy reward model provides nuanced incentives, encouraging more precise outputs. Experimental results demonstrate that GRPO with a standard 0/1 accuracy reward underperforms compared to supervised fine-tuning (SFT). In contrast, FGRPR, applied to Qwen2.5-VL(3B and 7B), surpasses all baseline models, including GPT4o, LLaMA2(90B), and SFT, across five in-domain datasets. On an out-of-domain dataset, FGRPR achieves performance comparable to SFT but excels when target values are larger, as its fuzzy reward function assigns higher rewards to closer approximations. This approach is broadly applicable to tasks where the precision of the answer is critical. Code and data: https://github.com/yeyimilk/CrowdVLM-R1

Yijiashun Qi, Shuzhang Cai, Zunduo Zhao et al.

As Artificial Intelligence (AI) has developed rapidly over the past few decades, the new generation of AI, Large Language Models (LLMs) trained on massive datasets, has achieved ground-breaking performance in many applications. Further progress has been made in multimodal LLMs, with many datasets created to evaluate LLMs with vision abilities. However, none of those datasets focuses solely on marine mammals, which are indispensable for ecological equilibrium. In this work, we build a benchmark dataset with 1,423 images of 65 kinds of marine mammals, where each animal is uniquely classified into different levels of class, ranging from species-level to medium-level to group-level. Moreover, we evaluate several approaches for classifying these marine mammals: (1) machine learning (ML) algorithms using embeddings provided by neural networks, (2) influential pre-trained neural networks, (3) zero-shot models: CLIP and LLMs, and (4) a novel LLM-based multi-agent system (MAS). The results demonstrate the strengths of traditional models and LLMs in different aspects, and the MAS can further improve the classification performance. The dataset is available on GitHub: https://github.com/yeyimilk/LLM-Vision-Marine-Animals.git.

Shuzhang Cai, Twumasi Mensah-Boateng, Xander Kuksov et al.

Large Language Models (LLMs) have demonstrated exceptional abilities across a broad range of language-related tasks, including generating solutions to complex reasoning problems. An effective technique to enhance LLM performance is in-context learning, which encourages a step-by-step reasoning process by including explanatory examples to guide the model's responses. However, selecting appropriate exemplars for the model poses a challenge, as each dataset demands a distinct set of exemplars to enable the LLM to learn effectively and perform well on the test set. Current studies often rely on uncertainty- or diversity-based selection strategies to select exemplars for annotation and to improve model learning. However, these studies typically employ a non-adaptive approach, selecting a set of exemplars all at once. We argue that this non-adaptive strategy may result in a set of exemplars with high redundancy in terms of the knowledge covered, ultimately reducing their overall informativeness. To address this limitation, we propose \textsc{Adaptive-Prompt}, a novel method that adaptively selects exemplars by leveraging model feedback from previously chosen exemplars. Experimental results show that \textsc{Adaptive-Prompt} significantly enhances LLM performance across a variety of reasoning tasks.

Shengjuan Cai, Fangxin Fang, Vincent-Henri Peuch et al.

PM2.5 forecasting is crucial for public health, air quality management, and policy development. Traditional physics-based models are computationally demanding and slow to adapt to real-time conditions. Deep learning models show potential in efficiency but still suffer from accuracy loss over time due to error accumulation. To address these challenges, we propose a dual deep neural network (D-DNet) prediction and data assimilation system that efficiently integrates real-time observations, ensuring reliable operational forecasting. D-DNet excels in global operational forecasting for PM2.5 and AOD550, maintaining consistent accuracy throughout the entire year of 2019. It demonstrates notably higher efficiency than the Copernicus Atmosphere Monitoring Service (CAMS) 4D-Var operational forecasting system while maintaining comparable accuracy. This efficiency benefits ensemble forecasting, uncertainty analysis, and large-scale tasks.