SEARCH · Searcharxiv

Results for “transformer architecture”

Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 865 records · Page 48Linked to original sources

arXiv2026

Multimodal embodiment-aware navigation transformer

Goal-conditioned navigation models for ground robots trained using supervised learning show promising zero-shot transfer, but their collision-avoidance capability nevertheless degrades under distribution shift, i.e. environmental, robot or sensor configuration changes. We propose ViLiNT a multimodal, attention-based policy for goal navigation, trained on heterogeneous data from multiple platforms and environments, which improves robustness with two key features. First, we fuse RGB images, 3D LiDAR point clouds, a goal embedding and a robot's embodiment descriptor with a transformer architecture to capture complementary geometry and appearance cues. The transformer's output is used to condition a diffusion model that generates navigable trajectories. Second, using automatically generated offline labels, we train a path clearance prediction head for scoring and ranking trajectories produced by the diffusion model. The diffusion conditioning as well as the trajectory ranking head depend on a robot's embodiment token that allows our model to generate and select trajectories with respect to the robot's dimensions. Across three simulated environments, ViLiNT improves Success Rate on average by 166\% over equivalent state-of-the-art vision-only baseline (NoMaD). This increase in performance is confirmed through real-world deployments of a rover navigating in obstacle fields. These results highlight that combining multimodal fusion with our collision prediction mechanism leads to improved off-road navigation robustness.

cs.RO↗

arXiv2026

GeoFormer: Geometry-Aware Transformer and its application to 5D First-Arrival Picking

We propose GeoFormer, a Geometry-Aware Transformer architecture specifically designed for prestack seismic data. Unlike Vision Transformer, whose tokens are extracted from 2D patches and primarily encode visual patterns, GeoFormer is designed for prestack seismic data by explicitly incorporating acquisition geometry. Each seismic trace is represented by a 5D unit consisting of the waveform and four source-receiver coordinates, from which two geometric attributes are derived: offset and relative elevation, where relative elevation is the receiver elevation minus the source elevation. GeoFormer therefore performs trace-level tokenization, where each token combines the waveform with these geometric attributes. To exploit these geometric attributes, GeoFormer introduces three geometry injection mechanisms operating at different levels of the Transformer pipeline. At the token level, GeomMLP replaces the classification token with a per-trace geometric representation derived from offset and elevation. At the normalization level, GeomAdaLN replaces uniform layer normalization with geometry-conditioned feature modulation. At the attention level, GeomAttnBias injects a parameter-free physical prior that geometrically proximate traces should attend more to each other. We validate GeoFormer on first-arrival picking, a representative seismic processing task that relies heavily on acquisition geometry. Experiments on four field datasets demonstrate that GeoFormer outperforms Vision Transformer and two task-specific baselines. Crucially, GeoFormer maintains robust picking accuracy under strong noise because its trace-level tokens encode geometric attributes that provide a physical prior independent of waveform quality. Ablation studies confirm that GeomMLP, GeomAdaLN, and GeomAttnBias each contribute to GeoFormer's picking accuracy.

physics.geo-ph↗

arXiv2026

Good Memory Has ECC: Evaluating the Memory of Vision-Language Models Beyond Accuracy

Memory is widely viewed as an important unsolved problem for LLMs and VLMs, and current benchmarks typically evaluate it by testing accuracy over long text or video. However, accuracy alone misses properties that matter for real long-horizon tasks. We introduce ECCBench, a benchmark and evaluation protocol that measures memory beyond a system's capacity--its raw accuracy at a specific budget--via three axes we call ECC: efficiency--the computation, in FLOPs, needed to answer from memory; compression--whether compressible inputs are remembered more accurately or efficiently; and calibration--whether the system abstains in response to its own uncertainty and the cost of an error. We find that pretrained VLMs compress their memory over text but not video and are poorly calibrated on both. Among a broader set of memory backbones, several non-Transformer architectures achieve better compression-calibration tradeoffs than RoPE Transformers, suggesting they may be useful components for agents operating over long horizons.

cs.LG↗

arXiv2026

WaveFoG: Wavelet Gated Transformer for Parkinson's Freezing of Gait Detection Using Wearable Accelerometer Signals

Freezing of gait (FoG) is one of the most debilitating episodic motor symptoms of Parkinson's disease (PD). FoG detection from wrist worn inertial signals poses challenging problems such as extreme data class imbalance (only 12-18% of signal windows are FoG), low frequency postural instabilities alongside diagnostically crucial 3-8 Hz rhythmic tremors, and the need for subject independent generalisation. To address these, this paper presents WaveFoG, a novel wavelet gated Transformer architecture that combines multi scale representations from discrete wavelet transform (DWT) with a dual branch encoder. A 1D CNN captures local gait texture, and a Transformer captures global temporal context. They are fused via a sigmoid gating fusion layer that modulates the encoder output based on the DWT sub-band descriptors. Trained with focal loss and evaluated on the public Kaggle TLVMC FoG Prediction dataset under subject independent grouped ten fold cross validation (SI-CV), WaveFoG achieves a mean F1-score of 0.875 +/- 0.017 and AUPRC of 0.833 +/- 0.020, exceeding single branch baselines (1D CNN, Bi-LSTM, and vanilla Transformer) by up to 5.4 pp on F1. Temporal saliency maps highlight temporal patterns that are consistent with reported characteristics of FoG onset. The source code of this project is available at: https://github.com/Shihabul-Shuvo/WaveFoG

eess.SP↗

arXiv2020

Evaluating the Impact of Knowledge Graph Context on Entity Disambiguation Models

Pretrained Transformer models have emerged as state-of-the-art approaches that learn contextual information from text to improve the performance of several NLP tasks. These models, albeit powerful, still require specialized knowledge in specific scenarios. In this paper, we argue that context derived from a knowledge graph (in our case: Wikidata) provides enough signals to inform pretrained transformer models and improve their performance for named entity disambiguation (NED) on Wikidata KG. We further hypothesize that our proposed KG context can be standardized for Wikipedia, and we evaluate the impact of KG context on state-of-the-art NED model for the Wikipedia knowledge base. Our empirical results validate that the proposed KG context can be generalized (for Wikipedia), and providing KG context in transformer architectures considerably outperforms the existing baselines, including the vanilla transformer models.

cs.CL↗

arXiv2025

Recent Advances in Transformer and Large Language Models for UAV Applications

The rapid advancement of Transformer-based models has reshaped the landscape of uncrewed aerial vehicle (UAV) systems by enhancing perception, decision-making, and autonomy. This review paper systematically categorizes and evaluates recent developments in Transformer architectures applied to UAVs, including attention mechanisms, CNN-Transformer hybrids, reinforcement learning Transformers, and large language models (LLMs). Unlike previous surveys, this work presents a unified taxonomy of Transformer-based UAV models, highlights emerging applications such as precision agriculture and autonomous navigation, and provides comparative analyses through structured tables and performance benchmarks. The paper also reviews key datasets, simulators, and evaluation metrics used in the field. Furthermore, it identifies existing gaps in the literature, outlines critical challenges in computational efficiency and real-time deployment, and offers future research directions. This comprehensive synthesis aims to guide researchers and practitioners in understanding and advancing Transformer-driven UAV technologies.

cs.CV↗

arXiv2025

Hybrid Quantum-Classical Ensemble Learning for S\&P 500 Directional Prediction

Financial market prediction is a challenging application of machine learning, where even small improvements in directional accuracy can yield substantial value. Most models struggle to exceed 55--57\% accuracy due to high noise, non-stationarity, and market efficiency. We introduce a hybrid ensemble framework combining quantum sentiment analysis, Decision Transformer architecture, and strategic model selection, achieving 60.14\% directional accuracy on S\&P 500 prediction, a 3.10\% improvement over individual models. Our framework addresses three limitations of prior approaches. First, architecture diversity dominates dataset diversity: combining different learning algorithms (LSTM, Decision Transformer, XGBoost, Random Forest, Logistic Regression) on the same data outperforms training identical architectures on multiple datasets (60.14\% vs.\ 52.80\%), confirmed by correlation analysis ($r>0.6$ among same-architecture models). Second, a 4-qubit variational quantum circuit enhances sentiment analysis, providing +0.8\% to +1.5\% gains per model. Third, smart filtering excludes weak predictors (accuracy $<52\%$), improving ensemble performance (Top-7 models: 60.14\% vs.\ all 35 models: 51.2\%). We evaluate on 2020--2023 market data across seven instruments, covering diverse regimes including the COVID-19 crash and inflation-driven correction. McNemar's test confirms statistical significance ($p<0.05$). Preliminary backtesting with confidence-based filtering (6+ model consensus) yields a Sharpe ratio of 1.2 versus buy-and-hold's 0.8, demonstrating practical trading potential.

cs.LG↗

arXiv2025

Scaling Up ESM2 Architectures for Long Protein Sequences Analysis: Long and Quantized Approaches

Various approaches utilizing Transformer architectures have achieved state-of-the-art results in Natural Language Processing (NLP). Based on this success, numerous architectures have been proposed for other types of data, such as in biology, particularly for protein sequences. Notably among these are the ESM2 architectures, pre-trained on billions of proteins, which form the basis of various state-of-the-art approaches in the field. However, the ESM2 architectures have a limitation regarding input size, restricting it to 1,022 amino acids, which necessitates the use of preprocessing techniques to handle sequences longer than this limit. In this paper, we present the long and quantized versions of the ESM2 architectures, doubling the input size limit to 2,048 amino acids.

cs.LG↗

arXiv2025

Beyond Black-Box Predictions: Identifying Marginal Feature Effects in Tabular Transformer Networks

In recent years, deep neural networks have showcased their predictive power across a variety of tasks. Beyond natural language processing, the transformer architecture has proven efficient in addressing tabular data problems and challenges the previously dominant gradient-based decision trees in these areas. However, this predictive power comes at the cost of intelligibility: Marginal feature effects are almost completely lost in the black-box nature of deep tabular transformer networks. Alternative architectures that use the additivity constraints of classical statistical regression models can maintain intelligible marginal feature effects, but often fall short in predictive power compared to their more complex counterparts. To bridge the gap between intelligibility and performance, we propose an adaptation of tabular transformer networks designed to identify marginal feature effects. We provide theoretical justifications that marginal feature effects can be accurately identified, and our ablation study demonstrates that the proposed model efficiently detects these effects, even amidst complex feature interactions. To demonstrate the model's predictive capabilities, we compare it to several interpretable as well as black-box models and find that it can match black-box performances while maintaining intelligibility. The source code is available at https://github.com/OpenTabular/NAMpy.

cs.LG↗

arXiv2017

Accelerating Discrete Wavelet Transforms on Parallel Architectures

The 2-D discrete wavelet transform (DWT) can be found in the heart of many image-processing algorithms. Until recently, several studies have compared the performance of such transform on various shared-memory parallel architectures, especially on graphics processing units (GPUs). All these studies, however, considered only separable calculation schemes. We show that corresponding separable parts can be merged into non-separable units, which halves the number of steps. In addition, we introduce an optional optimization approach leading to a reduction in the number of arithmetic operations. The discussed schemes were adapted on the OpenCL framework and pixel shaders, and then evaluated using GPUs of two biggest vendors. We demonstrate the performance of the proposed non-separable methods by comparison with existing separable schemes. The non-separable schemes outperform their separable counterparts on numerous setups, especially considering the pixel shaders.

cs.PF↗

arXiv2024

On the Performance and Memory Footprint of Distributed Training: An Empirical Study on Transformers

Transformer models have emerged as potent solutions to a wide array of multidisciplinary challenges. The deployment of Transformer architectures is significantly hindered by their extensive computational and memory requirements, necessitating the reliance on advanced efficient distributed training methodologies. Prior research has delved into the performance bottlenecks associated with distributed training, aiming to unravel these bottlenecks and suggest optimization directions. However, such analyses often overlook three aspects unique to Transformer models: the specialized architecture, the dependency on various distributed strategies, and the requirement to balance computational and memory overhead. This paper aims to bridge this gap by offering a comprehensive examination of the performance bottlenecks inherent in distributed training of Transformer models, leveraging both theoretical analysis and empirical investigation. We propose an analytical framework tailored to these unique aspects of Transformers, facilitating a holistic evaluation of model architectures, distributed strategies, and resource consumption. Based on this analytical framework, we conduct a comparative analysis of theoretical performances and further systematically explore how various distributed training strategies fare in real-world scenarios. Most of the experimental results can be well explained by the analytical outcomes derived from the analytical framework. Notably, our findings suggest an advantage of pipeline parallelism over data parallelism for Transformer models. Moreover, we shed light on some unexpected outcomes, such as the potential for increased total memory overhead due to suboptimal model partitioning within pipeline parallelism. Additionally, we underscore the significance of communication block size and waiting time to further enhance performance.

cs.DC↗

arXiv2023

1D-Convolutional transformer for Parkinson disease diagnosis from gait

This paper presents an efficient deep neural network model for diagnosing Parkinson's disease from gait. More specifically, we introduce a hybrid ConvNet-Transformer architecture to accurately diagnose the disease by detecting the severity stage. The proposed architecture exploits the strengths of both Convolutional Neural Networks and Transformers in a single end-to-end model, where the former is able to extract relevant local features from Vertical Ground Reaction Force (VGRF) signal, while the latter allows to capture long-term spatio-temporal dependencies in data. In this manner, our hybrid architecture achieves an improved performance compared to using either models individually. Our experimental results show that our approach is effective for detecting the different stages of Parkinson's disease from gait data, with a final accuracy of 88%, outperforming other state-of-the-art AI methods on the Physionet gait dataset. Moreover, our method can be generalized and adapted for other classification problems to jointly address the feature relevance and spatio-temporal dependency problems in 1D signals. Our source code and pre-trained models are publicly available at https://github.com/SafwenNaimi/1D-Convolutional-transformer-for-Parkinson-disease-diagnosis-from-gait.

cs.CV↗

arXiv2024

Neural Circuit Diagrams: Robust Diagrams for the Communication, Implementation, and Analysis of Deep Learning Architectures

Diagrams matter. Unfortunately, the deep learning community has no standard method for diagramming architectures. The current combination of linear algebra notation and ad-hoc diagrams fails to offer the necessary precision to understand architectures in all their detail. However, this detail is critical for faithful implementation, mathematical analysis, further innovation, and ethical assurances. I present neural circuit diagrams, a graphical language tailored to the needs of communicating deep learning architectures. Neural circuit diagrams naturally keep track of the changing arrangement of data, precisely show how operations are broadcast over axes, and display the critical parallel behavior of linear operations. A lingering issue with existing diagramming methods is the inability to simultaneously express the detail of axes and the free arrangement of data, which neural circuit diagrams solve. Their compositional structure is analogous to code, creating a close correspondence between diagrams and implementation. In this work, I introduce neural circuit diagrams for an audience of machine learning researchers. After introducing neural circuit diagrams, I cover a host of architectures to show their utility and breed familiarity. This includes the transformer architecture, convolution (and its difficult-to-explain extensions), residual networks, the U-Net, and the vision transformer. I include a Jupyter notebook that provides evidence for the close correspondence between diagrams and code. Finally, I examine backpropagation using neural circuit diagrams. I show their utility in providing mathematical insight and analyzing algorithms' time and space complexities.

cs.LG↗

arXiv2026

WISCA: A Lightweight Model Transition Method to Improve LLM Training via Weight Scaling

Transformer architecture gradually dominates the LLM field. Recent advances in training optimization for Transformer-based large language models (LLMs) primarily focus on architectural modifications or optimizer adjustments. However, these approaches lack systematic optimization of weight patterns during training. Weight pattern refers to the distribution and relative magnitudes of weight parameters in a neural network. To address this issue, we propose a Weight Scaling method called WISCA to enhance training efficiency and model quality by strategically improving neural network weight patterns without changing network structures. By rescaling weights while preserving model outputs, WISCA indirectly optimizes the model's training trajectory. Experiments demonstrate that WISCA significantly improves convergence quality (measured by generalization capability and loss reduction), particularly in LLMs with Grouped Query Attention (GQA) architectures and LoRA fine-tuning tasks. Empirical results show 5.6% average improvement on zero-shot validation tasks and 2.12% average reduction in training perplexity across multiple architectures.

cs.LG↗

arXiv2024

SmartMem: Layout Transformation Elimination and Adaptation for Efficient DNN Execution on Mobile

This work is motivated by recent developments in Deep Neural Networks, particularly the Transformer architectures underlying applications such as ChatGPT, and the need for performing inference on mobile devices. Focusing on emerging transformers (specifically the ones with computationally efficient Swin-like architectures) and large models (e.g., Stable Diffusion and LLMs) based on transformers, we observe that layout transformations between the computational operators cause a significant slowdown in these applications. This paper presents SmartMem, a comprehensive framework for eliminating most layout transformations, with the idea that multiple operators can use the same tensor layout through careful choice of layout and implementation of operations. Our approach is based on classifying the operators into four groups, and considering combinations of producer-consumer edges between the operators. We develop a set of methods for searching such layouts. Another component of our work is developing efficient memory layouts for 2.5 dimensional memory commonly seen in mobile devices. Our experimental results show that SmartMem outperforms 5 state-of-the-art DNN execution frameworks on mobile devices across 18 varied neural networks, including CNNs, Transformers with both local and global attention, as well as LLMs. In particular, compared to DNNFusion, SmartMem achieves an average speedup of 2.8$\times$, and outperforms TVM and MNN with speedups of 6.9$\times$ and 7.9$\times$, respectively, on average.

cs.LG↗

arXiv2025

CNN-TFT explained by SHAP with multi-head attention weights for time series forecasting

Convolutional neural networks (CNNs) and transformer architectures offer strengths for modeling temporal data: CNNs excel at capturing local patterns and translational invariances, while transformers effectively model long-range dependencies via self-attention. This paper proposes a hybrid architecture integrating convolutional feature extraction with a temporal fusion transformer (TFT) backbone to enhance multivariate time series forecasting. The CNN module first applies a hierarchy of one-dimensional convolutional layers to distill salient local patterns from raw input sequences, reducing noise and dimensionality. The resulting feature maps are then fed into the TFT, which applies multi-head attention to capture both short- and long-term dependencies and to weigh relevant covariates adaptively. We evaluate the CNN-TFT on a hydroelectric natural flow time series dataset. Experimental results demonstrate that CNN-TFT outperforms well-established deep learning models, with a mean absolute percentage error of up to 2.2%. The explainability of the model is obtained by a proposed Shapley additive explanations with multi-head attention weights (SHAP-MHAW). Our novel architecture, named CNN-TFT-SHAP-MHAW, is promising for applications requiring high-fidelity, multivariate time series forecasts, being available for future analysis at https://github.com/SFStefenon/CNN-TFT-SHAP-MHAW .

cs.LG↗

arXiv2025

Boosting Generative Adversarial Transferability with Self-supervised Vision Transformer Features

The ability of deep neural networks (DNNs) come from extracting and interpreting features from the data provided. By exploiting intermediate features in DNNs instead of relying on hard labels, we craft adversarial perturbation that generalize more effectively, boosting black-box transferability. These features ubiquitously come from supervised learning in previous work. Inspired by the exceptional synergy between self-supervised learning and the Transformer architecture, this paper explores whether exploiting self-supervised Vision Transformer (ViT) representations can improve adversarial transferability. We present dSVA -- a generative dual self-supervised ViT features attack, that exploits both global structural features from contrastive learning (CL) and local textural features from masked image modeling (MIM), the self-supervised learning paradigm duo for ViTs. We design a novel generative training framework that incorporates a generator to create black-box adversarial examples, and strategies to train the generator by exploiting joint features and the attention mechanism of self-supervised ViTs. Our findings show that CL and MIM enable ViTs to attend to distinct feature tendencies, which, when exploited in tandem, boast great adversarial generalizability. By disrupting dual deep features distilled by self-supervised ViTs, we are rewarded with remarkable black-box transferability to models of various architectures that outperform state-of-the-arts. Code available at https://github.com/spencerwooo/dSVA.

cs.CV↗

arXiv2021

PGT: Pseudo Relevance Feedback Using a Graph-Based Transformer

Most research on pseudo relevance feedback (PRF) has been done in vector space and probabilistic retrieval models. This paper shows that Transformer-based rerankers can also benefit from the extra context that PRF provides. It presents PGT, a graph-based Transformer that sparsifies attention between graph nodes to enable PRF while avoiding the high computational complexity of most Transformer architectures. Experiments show that PGT improves upon non-PRF Transformer reranker, and it is at least as accurate as Transformer PRF models that use full attention, but with lower computational costs.

cs.IR↗