13 papers
FrequencyFormer: A Co-Designed Sensor-to-Processor Pipeline for Frequency-Domain Vision Transformer Inference
Chengwei Zhou, Ovishake Sen, Xuming Chen +5
Deploying vision transformers (ViTs) on sensor-edge systems is limited not only by on-device compute, but also by the energy and bandwidth required to transmit high-dimensional ima…
Certifiable Safe RLHF: Semantic Grounding and Fixed Penalty Constraint Optimization for Safer LLM Alignment
Kartik Pandit, Sourav Ganguly, Arnesh Banerjee +2
Ensuring safety is a foundational requirement for large language models (LLMs). Achieving an appropriate balance between enhancing the utility of model outputs and mitigating their…
LIMCA: LLM for Automating Analog In-Memory Computing Architecture Design Exploration
Deepak Vungarala, Md Hasibul Amin, Pietro Mercati +5
Resistive crossbars enabling analog In-Memory Computing (IMC) have emerged as a promising architecture for Deep Neural Network (DNN) acceleration, offering high memory bandwidth an…
CacheTrap: Unveiling a Stealthier Gray-Box Trojan against LLMs
Mohaiminul Al Nahian, Abeer Matar A. Almalky, Gamana Aragonda +6
The rapid advancement of large language models (LLMs) has sparked growing interest in understanding their security vulnerabilities, particularly Trojan attacks that enable stealthy…
Light-Bound Transformers: Hardware-Anchored Robustness for Silicon-Photonic Computer Vision Systems
Xuming Chen, Deniz Najafi, Chengwei Zhou +6
Deploying Vision Transformers (ViTs) on near-sensor analog accelerators demands training pipelines that are explicitly aligned with device-level noise and energy constraints. We in…
GLANCE: Gaze-Led Attention Network for Compressed Edge-inference
Neeraj Solanki, Hong Ding, Sepehr Tabrizchi +4
Real-time object detection in AR/VR systems faces critical computational constraints, requiring sub-10\,ms latency within tight power budgets. Inspired by biological foveal vision,…