Architectures

The paper proposes SISA (SSM-Informed Softmax Attention), a novel hybrid attention mechanism that integrates state-space model (SSM) importance signals directly into the attention score, achieving sta…

Zamba2-VL is a new suite of vision-language models built on the Zamba2 hybrid architecture, achieving state-of-the-art performance and significantly improved inference efficiency compared to leading T…

This paper introduces Bifocal dLLMs (R2LM), a new paradigm for discrete diffusion language models that combines causal and bidirectional attention for improved throughput and generation quality.

AEGIS introduces a novel physics-based system that analyzes encrypted network traffic flow dynamics, achieving state-of-the-art zero-day evasion detection with high accuracy and low latency.

The paper demonstrates that the location and nature of state encoding in sequence models are not fixed architectural traits but are highly dependent on the specific task, showing that the encoding pro…

CaMBRAIN introduces a novel Mamba-based State Space Model (SSM) for real-time, continuous EEG inference, achieving state-of-the-art results with significantly higher throughput than existing methods.

This paper introduces IR275K, a curated benchmark for multi-frame super-resolution in infrared remote sensing, and evaluates CGMamba, a lightweight state-space model, achieving state-of-the-art perfor…

MARS proposes an encoder-agnostic aggregation operator that explicitly models multi-scale temporal structure in sequential recommendation, achieving state-of-the-art performance across both sparse and…

This paper introduces EVOTS, an evolutionary neural architecture search framework for discovering task-adaptive Transformer-like models for multivariate time-series forecasting.

The paper introduces a novel, transferable learned attack (LT-MIA) that detects a universal 'signature of memorization' in language models, achieving high accuracy across diverse model architectures (…

This paper provides the first systematic threat analysis of State-Space Models (SSMs) in safety-critical applications, introducing novel attack classes and formal metrics to quantify their security an…

This paper proposes DAPGNet, a dynamic adaptive physics-guided graph diffusion network for hyperspectral image classification, which achieves state-of-the-art performance on four datasets.

A regression-based approach is presented for Arabic dialect geolocation using a hierarchical neural architecture and spherical geodesic loss, achieving a median localization error of 481.2 km.

The paper proposes an SE ViT-BiLSTM hybrid model for enhanced intrusion detection in IIoT and IoMT environments, achieving superior performance on real-world datasets, especially after data balancing.

This paper proposes ConsisFormer, a compute-efficient Transformer design for wireless foundation models (WFMs) using short-term channel consistency and adaptive token aggregation.

SANA-Streaming introduces a novel, efficient framework that enables real-time, high-resolution streaming video-to-video editing by combining a hybrid diffusion transformer with specialized training an…

EigeNet introduces a geometry-informed multi-modal Transformer framework to achieve state-of-the-art few-shot novel view Room Impulse Response (RIR) prediction by effectively integrating spatial geome…

This paper argues for the importance of modularity and heterogeneity in AI architectures, contrasting the Transformer model with the structure of the cortex.

The paper demonstrates that in Mamba-2, single-bucket probes can detect a large functional signature (detection layer) that is not fully responsible for the actual computation (execution layer), chall…

This paper proposes and validates a novel hardware architecture, ITP-STDP, to significantly reduce the energy consumption and hardware overhead associated with training Spiking Neural Networks (SNNs).