Rickard Ewetz
YOU?
Author Swipe
View article: Knowledge Editing for Multi-Hop Question Answering Using Semantic Analysis
Knowledge Editing for Multi-Hop Question Answering Using Semantic Analysis Open
Large Language Models (LLMs) require lightweight avenues of updating stored information that has fallen out of date. Knowledge Editing (KE) approaches have been successful in updating model knowledge for simple factual queries but struggle…
View article: Verifiable Natural Language to Linear Temporal Logic Translation: A Benchmark Dataset and Evaluation Suite
Verifiable Natural Language to Linear Temporal Logic Translation: A Benchmark Dataset and Evaluation Suite Open
Empirical evaluation of state-of-the-art natural-language (NL) to temporal-logic (TL) translation systems reveals near-perfect performance on existing benchmarks. However, current studies measure only the accuracy of the translation of NL …
View article: PATCHOUT: Adversarial Patch Detection and Localization using Semantic Consistency
PATCHOUT: Adversarial Patch Detection and Localization using Semantic Consistency Open
Computer vision systems are actively deployed in safety-critical applications such as autonomous vehicles. Real-world adversarial patches are capable of compromising the artificial intelligence (AI) systems with catastrophic outcomes. Exis…
View article: LOGIC: Logic Synthesis for Digital In-Memory Computing
LOGIC: Logic Synthesis for Digital In-Memory Computing Open
In-memory processing offers a promising solution for enhancing the performance of data-intensive applications. While analog in-memory computing demonstrates remarkable efficiency, its limited precision is suitable only for approximate comp…
View article: NSP: A Neuro-Symbolic Natural Language Navigational Planner
NSP: A Neuro-Symbolic Natural Language Navigational Planner Open
Path planners that can interpret free-form natural language instructions hold promise to automate a wide range of robotics applications. These planners simplify user interactions and enable intuitive control over complex semi-autonomous sy…
View article: Data Augmentation for Image Classification using Generative AI
Data Augmentation for Image Classification using Generative AI Open
Scaling laws dictate that the performance of AI models is proportional to the amount of available data. Data augmentation is a promising solution to expanding the dataset size. Traditional approaches focused on augmentation using rotation,…
View article: Integrated Decision Gradients: Compute Your Attributions Where the Model Makes Its Decision
Integrated Decision Gradients: Compute Your Attributions Where the Model Makes Its Decision Open
Attribution algorithms are frequently employed to explain the decisions of neural network models. Integrated Gradients (IG) is an influential attribution method due to its strong axiomatic foundation. The algorithm is based on integrating …
View article: Neuro Symbolic Reasoning for Planning: Counterexample Guided Inductive Synthesis using Large Language Models and Satisfiability Solving
Neuro Symbolic Reasoning for Planning: Counterexample Guided Inductive Synthesis using Large Language Models and Satisfiability Solving Open
Generative large language models (LLMs) with instruct training such as GPT-4 can follow human-provided instruction prompts and generate human-like responses to these prompts. Apart from natural language responses, they have also been found…
View article: Neural Stochastic Differential Equations for Robust and Explainable Analysis of Electromagnetic Unintended Radiated Emissions
Neural Stochastic Differential Equations for Robust and Explainable Analysis of Electromagnetic Unintended Radiated Emissions Open
We present a comprehensive evaluation of the robustness and explainability of ResNet-like models in the context of Unintended Radiated Emission (URE) classification and suggest a new approach leveraging Neural Stochastic Differential Equat…
View article: Integrated Decision Gradients: Compute Your Attributions Where the Model Makes Its Decision
Integrated Decision Gradients: Compute Your Attributions Where the Model Makes Its Decision Open
Attribution algorithms are frequently employed to explain the decisions of neural network models. Integrated Gradients (IG) is an influential attribution method due to its strong axiomatic foundation. The algorithm is based on integrating …
View article: On the Robustness of AlphaFold: A COVID-19 Case Study
On the Robustness of AlphaFold: A COVID-19 Case Study Open
Protein folding neural networks (PFNNs) such as AlphaFold predict remarkably accurate structures of proteins compared to other approaches. However, the robustness of such networks has heretofore not been explored. This is particularly rele…
View article: ExplainIt!: A Tool for Computing Robust Attributions of DNNs
ExplainIt!: A Tool for Computing Robust Attributions of DNNs Open
Responsible integration of deep neural networks into the design of trustworthy systems requires the ability to explain decisions made by these models. Explainability and transparency are critical for system analysis, certification, and hum…
View article: Shaping Noise for Robust Attributions in Neural Stochastic Differential Equations
Shaping Noise for Robust Attributions in Neural Stochastic Differential Equations Open
Neural SDEs with Brownian motion as noise lead to smoother attributions than traditional ResNets. Various attribution methods such as saliency maps, integrated gradients, DeepSHAP and DeepLIFT have been shown to be more robust for neural S…
View article: Protein Folding Neural Networks Are Not Robust
Protein Folding Neural Networks Are Not Robust Open
Deep neural networks such as AlphaFold and RoseTTAFold predict remarkably accurate structures of proteins compared to other algorithmic approaches. It is known that biologically small perturbations in the protein sequence do not lead to dr…
View article: On Smoother Attributions using Neural Stochastic Differential Equations
On Smoother Attributions using Neural Stochastic Differential Equations Open
Several methods have recently been developed for computing attributions of a neural network's prediction over the input features. However, these existing approaches for computing attributions are noisy and not robust to small perturbations…
View article: An Extension of Fano's Inequality for Characterizing Model Susceptibility to Membership Inference Attacks
An Extension of Fano's Inequality for Characterizing Model Susceptibility to Membership Inference Attacks Open
Deep neural networks have been shown to be vulnerable to membership inference attacks wherein the attacker aims to detect whether specific input data were used to train the model. These attacks can potentially leak private or proprietary d…
View article: An Extension of Fano's Inequality for Characterizing Model\n Susceptibility to Membership Inference Attacks
An Extension of Fano's Inequality for Characterizing Model\n Susceptibility to Membership Inference Attacks Open
Deep neural networks have been shown to be vulnerable to membership inference\nattacks wherein the attacker aims to detect whether specific input data were\nused to train the model. These attacks can potentially leak private or\nproprietar…
View article: Synthesis of Clock Networks with a Mode Reconfigurable Topology and No Short Circuit Current
Synthesis of Clock Networks with a Mode Reconfigurable Topology and No Short Circuit Current Open
Circuits deployed in the Internet of Things operate in low and high performance modes to cater to variable frequency and power requirements. Consequently, the clock networks for such circuits must be synthesized meeting drastically differe…
View article: Representable Matrices: Enabling High Accuracy Analog Computation for Inference of DNNs using Memristors
Representable Matrices: Enabling High Accuracy Analog Computation for Inference of DNNs using Memristors Open
Analog computing based on memristor technology is a promising solution to accelerating the inference phase of deep neural networks (DNNs). A fundamental problem is to map an arbitrary matrix to a memristor crossbar array (MCA) while maximi…
View article: Noise Injection Adaption
Noise Injection Adaption Open
In this work, we investigate various non-ideal effects (Stuck-At-Fault (SAF), IR-drop, thermal noise, shot noise, and random telegraph noise)of ReRAM crossbar when employing it as a dot-product engine for deep neural network (DNN) accelera…
View article: Clock Tree Construction based on Arrival Time Constraints
Clock Tree Construction based on Arrival Time Constraints Open
There are striking differences between constructing clock trees based on dynamic implied skew constraints and based on static arrival time constraints. Dynamic implied skew constraints allow the full timing margins to be utilized, but the …
View article: Construction of Latency-Bounded Clock Trees
Construction of Latency-Bounded Clock Trees Open
Clock trees must be constructed to function even under the influence of on-chip variations (OCV). Bounding the latency of a clock tree, i.e., the maximum delay from the tree root to any sequential element, is important because the latency …
View article: Synthesis of Clock Trees with Useful Skew based on Sparse-Graph Algorithms
Synthesis of Clock Trees with Useful Skew based on Sparse-Graph Algorithms Open
In this thesis, an optimization framework is proposed to synthesize clock trees with useful skews. The useful skews facilitate both low resource utilization and robustness to on-chip variations (OCV). First, techniques are proposed to cons…