Jundong Xu, Hao Fei, Meng Luo, Qian Liu, Liangming Pan, William Yang Wang, Preslav Nakov, Mong-Li Lee, Wynne Hsu. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

The design of loss function is of significant importance to optimize the quality of image transmission by using deep learning such as Joint Source-channel Coding (JSCC), especially with different Signal-to-Noise Ratio (SNR) conditions. Traditional methods using the fixed loss function lead to quality degradation when the real SNR is different from the SNR condition of the training environment. In this paper, an adaptive loss function based on SNR regulation is proposed. By combining Mean Square Error (MSE) and Structural Similarity Index Measure (SSIM), the weight of the loss function is automatically adjusted according to the changes of channel SNR. We show how to optimize the structural fidelity and visual quality of images in different SNR environments by adjusting the weight ratio. The experimental results show that the proposed method can effectively improve the image quality under low SNR conditions, while maintaining the accuracy of image details under high SNR conditions.

Multimodal symbolic logical reasoning, which aims to deduce new facts from multimodal input via formal logic, is critical in high-stakes applications such as autonomous driving and medical diagnosis, as its rigorous, deterministic reasoning helps prevent serious consequences. To evaluate such capabilities of current state-of-the-art vision language models (VLMs), we introduce the first benchmark MuSLR for multimodal symbolic logical reasoning grounded in formal logical rules. MuSLR comprises 1,093 instances across 7 domains, including 35 atomic symbolic logic and 976 logical combinations, with reasoning depths ranging from 2 to 9. We evaluate 7 state-of-the-art VLMs on MuSLR and find that they all struggle with multimodal symbolic reasoning, with the best model, GPT-4.1, achieving only 46.8%. Thus, we propose LogiCAM, a modular framework that applies formal logical rules to multimodal inputs, boosting GPT-4.1’s Chain-of-Thought performance by 14.13%, and delivering even larger gains on complex logics such as first-order logic. We also conduct a comprehensive error analysis, showing that around 70% of failures stem from logical misalignment between modalities, offering key insights to guide future improvements.

While the recent Chain-of-Thought (CoT) technique enhances the reasoning ability of large language models (LLMs) with the theory of mind, it might still struggle in handling logical reasoning that relies much on symbolic expressions and rigid deducing rules.To strengthen the logical reasoning capability of LLMs, we propose a novel Symbolic Chain-of-Thought, namely SymbCoT, a fully LLM-based framework that integrates symbolic expressions and logic rules with CoT prompting.Technically, building upon an LLM, SymbCoT 1) first translates the natural language context into the symbolic format, and then 2) derives a step-by-step plan to solve the problem with symbolic logical rules, 3) followed by a verifier to check the translation and reasoning chain.Via thorough evaluations on 5 standard datasets with both First-Order Logic and Constraint Optimization symbolic expressions, SymbCoT shows striking improvements over the CoT method consistently, meanwhile refreshing the current stateof-the-art performances.We further demonstrate that our system advances in more faithful, flexible, and explainable logical reasoning.To our knowledge, this is the first to combine symbolic expressions and rules into CoT for logical reasoning with LLMs.Code is open at https://github.com/Aiden0526/SymbCoT.