Provable Failure of Language Models in Learning Majority Boolean Logic via Gradient Descent

Recent advancements in Transformer-based architectures have led to impressive breakthroughs in natural language processing tasks, with models such as GPT-4, Claude, and Gemini demonstrating human-level reasoning abilities. However, despite their high performance, concerns remain about the inherent limitations of these models, especially when it comes to learning basic logical functions. While complexity-theoretic analyses indicate that Transformers can represent simple logic functions (e.g., $\mathsf{AND}$, $\mathsf{OR}$, and majority gates) by its nature of belonging to the $\mathsf{TC}^0$ class, these results assume ideal parameter settings and do not account for the constraints imposed by gradient descent-based training methods. In this work, we investigate whether Transformers can truly learn simple majority functions when trained using gradient-based methods. We focus on a simplified variant of the Transformer architecture and consider both $n=\mathrm{poly}(d)$ and $n=\exp(\Omega(d))$ number of training samples, where each sample is a $d$-size binary string paired with the output of a basic majority function. Our analysis demonstrates that even after $\mathrm{poly}(d)$ gradient queries, the generalization error of the Transformer model still remains substantially large, growing exponentially with $d$. This work highlights fundamental optimization challenges in training Transformers for the simplest logical reasoning tasks and provides new insights into their theoretical limitations.

Can You Count to Nine? A Human Evaluation Benchmark for Counting Limits in Modern Text-to-Video Models

Generative models have driven significant progress in a variety of AI tasks, including text-to-video generation, where models like Video LDM and Stable Video Diffusion can produce realistic, movie-level videos from textual instructions. Despite these advances, current text-to-video models still face fundamental challenges in reliably following human commands, particularly in adhering to simple numerical constraints. In this work, we present T2VCountBench, a specialized benchmark aiming at evaluating the counting capability of SOTA text-to-video models as of 2025. Our benchmark employs rigorous human evaluations to measure the number of generated objects and covers a diverse range of generators, covering both open-source and commercial models. Extensive experiments reveal that all existing models struggle with basic numerical tasks, almost always failing to generate videos with an object count of 9 or fewer. Furthermore, our comprehensive ablation studies explore how factors like video style, temporal dynamics, and multilingual inputs may influence counting performance. We also explore prompt refinement techniques and demonstrate that decomposing the task into smaller subtasks does not easily alleviate these limitations. Our findings highlight important challenges in current text-to-video generation and provide insights for future research aimed at improving adherence to basic numerical constraints.

MEET: A Million-Scale Dataset for Fine-Grained Geospatial Scene Classification with Zoom-Free Remote Sensing Imagery

Accurate fine-grained geospatial scene classification using remote sensing imagery is essential for a wide range of applications. However, existing approaches often rely on manually zooming remote sensing images at different scales to create typical scene samples. This approach fails to adequately support the fixed-resolution image interpretation requirements in real-world scenarios. To address this limitation, we introduce the Million-scale finE-grained geospatial scEne classification dataseT (MEET), which contains over 1.03 million zoom-free remote sensing scene samples, manually annotated into 80 fine-grained categories. In MEET, each scene sample follows a scene-inscene layout, where the central scene serves as the reference, and auxiliary scenes provide crucial spatial context for finegrained classification. Moreover, to tackle the emerging challenge of scene-in-scene classification, we present the Context-Aware Transformer (CAT), a model specifically designed for this task, which adaptively fuses spatial context to accurately classify the scene samples. CAT adaptively fuses spatial context to accurately classify the scene samples by learning attentional features that capture the relationships between the center and auxiliary scenes. Based on MEET, we establish a comprehensive benchmark for fine-grained geospatial scene classification, evaluating CAT against 11 competitive baselines. The results demonstrate that CAT significantly outperforms these baselines, achieving a 1.88% higher balanced accuracy (BA) with the Swin-Large backbone, and a notable 7.87% improvement with the Swin-Huge backbone. Further experiments validate the effectiveness of each module in CAT and show the practical applicability of CAT in the urban functional zone mapping. The source code and dataset will be publicly available at https://jerrywyn.github.io/project/MEET.html.

Text-to-Image Diffusion Models Cannot Count, and Prompt Refinement Cannot Help

Generative modeling is widely regarded as one of the most essential problems in today's AI community, with text-to-image generation having gained unprecedented real-world impacts. Among various approaches, diffusion models have achieved remarkable success and have become the de facto solution for text-to-image generation. However, despite their impressive performance, these models exhibit fundamental limitations in adhering to numerical constraints in user instructions, frequently generating images with an incorrect number of objects. While several prior works have mentioned this issue, a comprehensive and rigorous evaluation of this limitation remains lacking. To address this gap, we introduce T2ICountBench, a novel benchmark designed to rigorously evaluate the counting ability of state-of-the-art text-to-image diffusion models. Our benchmark encompasses a diverse set of generative models, including both open-source and private systems. It explicitly isolates counting performance from other capabilities, provides structured difficulty levels, and incorporates human evaluations to ensure high reliability. Extensive evaluations with T2ICountBench reveal that all state-of-the-art diffusion models fail to generate the correct number of objects, with accuracy dropping significantly as the number of objects increases. Additionally, an exploratory study on prompt refinement demonstrates that such simple interventions generally do not improve counting accuracy. Our findings highlight the inherent challenges in numerical understanding within diffusion models and point to promising directions for future improvements.

Integrated Computation and Communication with Fiber-optic Transmissions

Fiber-optic transmission systems are leveraged not only as high-speed communication channels but also as nonlinear kernel functions for machine learning computations, enabling the seamless integration of computational intelligence and communication.

Indiana Jones: There Are Always Some Useful Ancient Relics

This paper introduces Indiana Jones, an innovative approach to jailbreaking Large Language Models (LLMs) by leveraging inter-model dialogues and keyword-driven prompts. Through orchestrating interactions among three specialised LLMs, the method achieves near-perfect success rates in bypassing content safeguards in both white-box and black-box LLMs. The research exposes systemic vulnerabilities within contemporary models, particularly their susceptibility to producing harmful or unethical outputs when guided by ostensibly innocuous prompts framed in historical or contextual contexts. Experimental evaluations highlight the efficacy and adaptability of Indiana Jones, demonstrating its superiority over existing jailbreak methods. These findings emphasise the urgent need for enhanced ethical safeguards and robust security measures in the development of LLMs. Moreover, this work provides a critical foundation for future studies aimed at fortifying LLMs against adversarial exploitation while preserving their utility and flexibility.

MV-GMN: State Space Model for Multi-View Action Recognition

Recent advancements in multi-view action recognition have largely relied on Transformer-based models. While effective and adaptable, these models often require substantial computational resources, especially in scenarios with multiple views and multiple temporal sequences. Addressing this limitation, this paper introduces the MV-GMN model, a state-space model specifically designed to efficiently aggregate multi-modal data (RGB and skeleton), multi-view perspectives, and multi-temporal information for action recognition with reduced computational complexity. The MV-GMN model employs an innovative Multi-View Graph Mamba network comprising a series of MV-GMN blocks. Each block includes a proposed Bidirectional State Space Block and a GCN module. The Bidirectional State Space Block introduces four scanning strategies, including view-prioritized and time-prioritized approaches. The GCN module leverages rule-based and KNN-based methods to construct the graph network, effectively integrating features from different viewpoints and temporal instances. Demonstrating its efficacy, MV-GMN outperforms the state-of-the-arts on several datasets, achieving notable accuracies of 97.3\% and 96.7\% on the NTU RGB+D 120 dataset in cross-subject and cross-view scenarios, respectively. MV-GMN also surpasses Transformer-based baselines while requiring only linear inference complexity, underscoring the model's ability to reduce computational load and enhance the scalability and applicability of multi-view action recognition technologies.

ADKGD: Anomaly Detection in Knowledge Graphs with Dual-Channel Training

In the current development of large language models (LLMs), it is important to ensure the accuracy and reliability of the underlying data sources. LLMs are critical for various applications, but they often suffer from hallucinations and inaccuracies due to knowledge gaps in the training data. Knowledge graphs (KGs), as a powerful structural tool, could serve as a vital external information source to mitigate the aforementioned issues. By providing a structured and comprehensive understanding of real-world data, KGs enhance the performance and reliability of LLMs. However, it is common that errors exist in KGs while extracting triplets from unstructured data to construct KGs. This could lead to degraded performance in downstream tasks such as question-answering and recommender systems. Therefore, anomaly detection in KGs is essential to identify and correct these errors. This paper presents an anomaly detection algorithm in knowledge graphs with dual-channel learning (ADKGD). ADKGD leverages a dual-channel learning approach to enhance representation learning from both the entity-view and triplet-view perspectives. Furthermore, using a cross-layer approach, our framework integrates internal information aggregation and context information aggregation. We introduce a kullback-leibler (KL)-loss component to improve the accuracy of the scoring function between the dual channels. To evaluate ADKGD's performance, we conduct empirical studies on three real-world KGs: WN18RR, FB15K, and NELL-995. Experimental results demonstrate that ADKGD outperforms the state-of-the-art anomaly detection algorithms. The source code and datasets are publicly available at https://github.com/csjywu1/ADKGD.

On the Computational Capability of Graph Neural Networks: A Circuit Complexity Bound Perspective

Graph Neural Networks (GNNs) have become the standard approach for learning and reasoning over relational data, leveraging the message-passing mechanism that iteratively propagates node embeddings through graph structures. While GNNs have achieved significant empirical success, their theoretical limitations remain an active area of research. Existing studies primarily focus on characterizing GNN expressiveness through Weisfeiler-Lehman (WL) graph isomorphism tests. In this paper, we take a fundamentally different approach by exploring the computational limitations of GNNs through the lens of circuit complexity. Specifically, we analyze the circuit complexity of common GNN architectures and prove that under constraints of constant-depth layers, linear or sublinear embedding sizes, and polynomial precision, GNNs cannot solve key problems such as graph connectivity and graph isomorphism unless $\mathsf{TC}^0 = \mathsf{NC}^1$. These results reveal the intrinsic expressivity limitations of GNNs behind their empirical success and introduce a novel framework for analyzing GNN expressiveness that can be extended to a broader range of GNN models and graph decision problems.

Distilling Desired Comments for Enhanced Code Review with Large Language Models

There has been a growing interest in using Large Language Models (LLMs) for code review thanks to their proven proficiency in code comprehension. The primary objective of most review scenarios is to generate desired review comments (DRCs) that explicitly identify issues to trigger code fixes. However, existing LLM-based solutions are not so effective in generating DRCs for various reasons such as hallucination. To enhance their code review ability, they need to be fine-tuned with a customized dataset that is ideally full of DRCs. Nevertheless, such a dataset is not yet available, while manual annotation of DRCs is too laborious to be practical. In this paper, we propose a dataset distillation method, Desiview, which can automatically construct a distilled dataset by identifying DRCs from a code review dataset. Experiments on the CodeReviewer dataset comprising more than 150K review entries show that Desiview achieves an impressive performance of 88.93%, 80.37%, 86.67%, and 84.44% in terms of Precision, Recall, Accuracy, and F1, respectively, surpassing state-of-the-art methods. To validate the effect of such a distilled dataset on enhancing LLMs' code review ability, we first fine-tune the latest LLaMA series (i.e., LLaMA 3 and LLaMA 3.1) to build model Desiview4FT. We then enhance the model training effect through KTO alignment by feeding those review comments identified as non-DRCs to the LLMs, resulting in model Desiview4FA. Verification results indicate that Desiview4FA slightly outperforms Desiview4FT, while both models have significantly improved against the base models in terms of generating DRCs. Human evaluation confirms that both models identify issues more accurately and tend to generate review comments that better describe the issues contained in the code than the base LLMs do.