2 citations · 3 across the 14 of their papers we have counts for
17 papers
When Models Know but Can't Linearly Say: Relational Encoding Limits Linear Vulnerability Classification in Code LLMs
Rui Melo, Andre Catarino, Claudia Mamede +2
Large language models can detect vulnerable code, yet the form of vulnerability encoding in their representations remains unclear. We analyze 8 code models spanning 6.7B-15B parame…
SeFaR: Semantic Feature-aware Robustness Testing of Deep Neural Networks
Nusrat Jahan Mozumder, Divya Gopinath, Corina Pasareanu +1
Deep neural networks are increasingly deployed in safety-critical domains as perception modules, where failures are often caused due to rare and under-represented scenarios. This n…
Concept-Based Abductive and Contrastive Explanations for Behaviors of Vision Models
Ronaldo Canizales, Divya Gopinath, Corina Păsăreanu +1
*Concept-based explanations* offer a promising approach for explaining the predictions of deep neural networks in terms of high-level, human-understandable concepts. However, exist…
SecCodePRM: A Process Reward Model for Code Security
Weichen Yu, Ravi Mangal, Yinyi Luo +4
Large Language Models are rapidly becoming core components of modern software development workflows, yet ensuring code security remains challenging. Existing vulnerability detectio…
Fighting AI with AI: Leveraging Foundation Models for Assuring AI-Enabled Safety-Critical Systems
Anastasia Mavridou, Divya Gopinath, Corina S. Păsăreanu
The integration of AI components, particularly Deep Neural Networks (DNNs), into safety-critical systems such as aerospace and autonomous vehicles presents fundamental challenges f…
When "Correct" Is Not Safe: Can We Trust Functionally Correct Patches Generated by Code Agents?
Yibo Peng, James Song, Lei Li +6
Code agents are increasingly trusted to autonomously fix bugs on platforms such as GitHub, yet their security evaluation focuses almost exclusively on functional correctness. In th…
On the Difficulty of Selecting Few-Shot Examples for Effective LLM-based Vulnerability Detection
Md Abdul Hannan, Ronghao Ni, Chi Zhang +3
Large language models (LLMs) have demonstrated impressive capabilities across a wide range of coding tasks, including summarization, translation, completion, and code generation. D…
Prophecy: Inferring Formal Properties from Neuron Activations
Divya Gopinath, Corina S. Pasareanu, Muhammad Usman
We present Prophecy, a tool for automatically inferring formal properties of feed-forward neural networks. Prophecy is based on the observation that a significant part of the logic…
Conformal Safety Shielding for Imperfect-Perception Agents
William Scarbro, Calum Imrie, Sinem Getir Yaman +4
We consider the problem of safe control in discrete autonomous agents that use learned components for imperfect perception (or more generally, state estimation) from high-dimension…
A Mixture of Linear Corrections Generates Secure Code
Weichen Yu, Ravi Mangal, Terry Zhuo +2
Large language models (LLMs) have become proficient at sophisticated code-generation tasks, yet remain ineffective at reliably detecting or avoiding code vulnerabilities. Does this…
Report on NSF Workshop on Science of Safe AI
Rajeev Alur, Greg Durrett, Hadas Kress-Gazit +2
Recent advances in machine learning, particularly the emergence of foundation models, are leading to new opportunities to develop technology-based solutions to societal problems. H…
Scenario-based Compositional Verification of Autonomous Systems with Neural Perception
Christopher Watson, Rajeev Alur, Divya Gopinath +2
Recent advances in deep learning have enabled the development of autonomous systems that use deep neural networks for perception. Formal verification of these systems is challengin…
Debugging and Runtime Analysis of Neural Networks with VLMs (A Case Study)
Boyue Caroline Hu, Divya Gopinath, Corina S. Pasareanu +3
Debugging of Deep Neural Networks (DNNs), particularly vision models, is very challenging due to the complex and opaque decision-making processes in these networks. In this paper,…
Validating Mechanistic Interpretations: An Axiomatic Approach
Nils Palumbo, Ravi Mangal, Zifan Wang +3
Mechanistic interpretability aims to reverse engineer the computation performed by a neural network in terms of its internal components. Although there is a growing body of researc…
Concept-based Analysis of Neural Networks via Vision-Language Models
Ravi Mangal, Nina Narodytska, Divya Gopinath +4
The analysis of vision-based deep neural networks (DNNs) is highly desirable but it is very challenging due to the difficulty of expressing formal specifications for vision tasks a…
Transfer Attacks and Defenses for Large Language Models on Coding Tasks
Chi Zhang, Zifan Wang, Ravi Mangal +3
Modern large language models (LLMs), such as ChatGPT, have demonstrated impressive capabilities for coding tasks including writing and reasoning about code. They improve upon previ…
Is Certifying Robustness Still Worthwhile?
Ravi Mangal, Klas Leino, Zifan Wang +5
Over the years, researchers have developed myriad attacks that exploit the ubiquity of adversarial examples, as well as defenses that aim to guard against the security vulnerabilit…