1 citations · 2 across the 7 of their papers we have counts for
7 papers
Breaking Bad: Interpretability-Based Safety Audits of State-of-the-Art LLMs
Krishiv Agarwal, Ramneet Kaur, Colin Samplawski +6
Effective safety auditing of large language models (LLMs) demands tools that go beyond black-box probing and systematically uncover vulnerabilities rooted in model internals. We pr…
Jailbreaking the Matrix: Nullspace Steering for Controlled Model Subversion
Vishal Pramanik, Maisha Maliha, Susmit Jha +1
Large language models remain vulnerable to jailbreak attacks -- inputs designed to bypass safety mechanisms and elicit harmful responses -- despite advances in alignment and instru…
Spatio-Temporal Pruning for Compressed Spiking Large Language Models
Yi Jiang, Malyaban Bal, Brian Matejek +3
Large Language Models (LLMs) present significant challenges for deployment in energy-constrained environments due to their large model sizes and high inference latency. Spiking Neu…
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,…
Second-Order Forward-Mode Automatic Differentiation for Optimization
Adam D. Cobb, Atılım Güneş Baydin, Barak A. Pearlmutter +1
This paper introduces a second-order hyperplane search, a novel optimization step that generalizes a second-order line search from a line to a -dimensional hyperplane. This, com…
Task-Agnostic Detector for Insertion-Based Backdoor Attacks
Weimin Lyu, Xiao Lin, Songzhu Zheng +4
Textual backdoor attacks pose significant security threats. Current detection approaches, typically relying on intermediate feature representation or reconstructing potential trigg…
Calibration and Correctness of Language Models for Code
Claudio Spiess, David Gros, Kunal Suresh Pai +6
Machine learning models are widely used, but can also often be wrong. Users would benefit from a reliable indication of whether a given output from a given model should be trusted,…