Mansi Phute
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mansi Phute has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Alec Helbling
Georgia Institute of Technology
4 shared publications - Duen Horng Chau
Georgia Institute of Technology, Atlanta, GA, USA
3 shared publications - ShengYun Peng
Georgia Tech Georgia Tech Georgia Tech Georgia Tech Intel Labs Intel Labs Georgia Tech
3 shared publications Duen Horng Chau
Georgia Institute of Technology
2 shared publicationsMatthew Hull
Georgia Institute of Technology
2 shared publications- Minsuk Kahng
Google research
2 shared publications
Latest Publications
- 5 October 2026
- Published by Springer Nature
- p. 495-510
VISOR++ : Visual Input Based Steering for Large Vision Language Models
- 5 October 2026
- Published by Springer Nature
- p. 495-510
VISOR++ : Visual Input Based Steering for Large Vision Language Models
- 9 June 2026
- Published by Springer Nature
- p. 107-120
Differentiable Rendering Powered End-to-End Adversarial Attack Evaluation
- 9 June 2026
- Published by Springer Nature
- p. 107-120
Differentiable Rendering Powered End-to-End Adversarial Attack Evaluation
- Open Access
- 1 September 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 12, 10473-10480
RenderBender: A Survey on Adversarial Attacks Using Differentiable Rendering
Differentiable rendering techniques like Gaussian Splatting and Neural Radiance Fields have become powerful tools for generating high-fidelity models of 3D objects and scenes. Their ability to produce both physically plausible and differentiable models of scenes are key ingredient needed to produce physically plausible adversarial attacks on DNNs. However, the adversarial machine learning community has yet to fully explore these capabilities, partly due to differing attack goals (e.g., misclassification, misdetection) and a wide range of possible scene manipulations used to achieve them (e.g., alter texture, mesh). This survey contributes a framework that unifies diverse goals and tasks, facilitating easy comparison of existing work, identifying research gaps, and highlighting future directions—ranging from expanding attack goals and tasks to account for new modalities, state-of-the-art models, tools, and pipelines, to underscoring the importance of studying real-world threats in complex scenes.
- 1 September 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 12, 10473-10480
Open AccessRenderBender: A Survey on Adversarial Attacks Using Differentiable Rendering
Differentiable rendering techniques like Gaussian Splatting and Neural Radiance Fields have become powerful tools for generating high-fidelity models of 3D objects and scenes. Their ability to produce both physically plausible and differentiable models of scenes are key ingredient needed to produce physically plausible adversarial attacks on DNNs. However, the adversarial machine learning community has yet to fully explore these capabilities, partly due to differing attack goals (e.g., misclassification, misdetection) and a wide range of possible scene manipulations used to achieve them (e.g., alter texture, mesh). This survey contributes a framework that unifies diverse goals and tasks, facilitating easy comparison of existing work, identifying research gaps, and highlighting future directions—ranging from expanding attack goals and tasks to account for new modalities, state-of-the-art models, tools, and pipelines, to underscoring the importance of studying real-world threats in complex scenes.
- Open Access
- 11 April 2025
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 39 (28) , 29655-29657
LLM Attributor: Interactive Visual Attribution for LLM Generation
While large language models (LLMs) have shown remarkable capability to generate convincing text across diverse domains, concerns around its potential risks have highlighted the importance of understanding the rationale behind text generation. We present LLM ATTRIBUTOR, a Python library that provides interactive visualizations for training data attribution of an LLM’s text generation. Our library offers a new way to quickly attribute an LLM’s text generation to training data points to inspect model behaviors, enhance its trustworthiness, and compare model-generated text with user-provided text. Thanks to LLM ATTRIBUTOR’s broad support for computational notebooks, users can easily integrate it into their workflow to interactively visualize attributions of their models.
- 11 April 2025
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 39 (28) , 29655-29657
Open AccessLLM Attributor: Interactive Visual Attribution for LLM Generation
While large language models (LLMs) have shown remarkable capability to generate convincing text across diverse domains, concerns around its potential risks have highlighted the importance of understanding the rationale behind text generation. We present LLM ATTRIBUTOR, a Python library that provides interactive visualizations for training data attribution of an LLM’s text generation. Our library offers a new way to quickly attribute an LLM’s text generation to training data points to inspect model behaviors, enhance its trustworthiness, and compare model-generated text with user-provided text. Thanks to LLM ATTRIBUTOR’s broad support for computational notebooks, users can easily integrate it into their workflow to interactively visualize attributions of their models.
- Open Access
LLM Self Defense: By Self Examination, LLMs Know They Are Being Tricked
Large language models (LLMs) are popular for high-quality text generation but can produce harmful content, even when aligned with human values through reinforcement learning. Adversarial prompts can bypass their safety measures. We propose LLM Self Defense, a simple approach to defend against these attacks by having an LLM screen the induced responses. Our method does not require any fine-tuning, input preprocessing, or iterative output generation. Instead, we incorporate the generated content into a pre-defined prompt and employ another instance of an LLM to analyze the text and predict whether it is harmful. We test LLM Self Defense on GPT 3.5 and Llama 2, two of the current most prominent LLMs against various types of attacks, such as forcefully inducing affirmative responses to prompts and prompt engineering attacks. Notably, LLM Self Defense succeeds in reducing the attack success rate to virtually 0 using both GPT 3.5 and Llama 2. The code is publicly available at https://github.com/poloclub/llm-self-defense
Open AccessLLM Self Defense: By Self Examination, LLMs Know They Are Being Tricked
Large language models (LLMs) are popular for high-quality text generation but can produce harmful content, even when aligned with human values through reinforcement learning. Adversarial prompts can bypass their safety measures. We propose LLM Self Defense, a simple approach to defend against these attacks by having an LLM screen the induced responses. Our method does not require any fine-tuning, input preprocessing, or iterative output generation. Instead, we incorporate the generated content into a pre-defined prompt and employ another instance of an LLM to analyze the text and predict whether it is harmful. We test LLM Self Defense on GPT 3.5 and Llama 2, two of the current most prominent LLMs against various types of attacks, such as forcefully inducing affirmative responses to prompts and prompt engineering attacks. Notably, LLM Self Defense succeeds in reducing the attack success rate to virtually 0 using both GPT 3.5 and Llama 2. The code is publicly available at https://github.com/poloclub/llm-self-defense
- Open Access
LLM Attributor: Interactive Visual Attribution for LLM Generation
While large language models (LLMs) have shown remarkable capability to generate convincing text across diverse domains, concerns around its potential risks have highlighted the importance of understanding the rationale behind text generation. We present LLM Attributor, a Python library that provides interactive visualizations for training data attribution of an LLM's text generation. Our library offers a new way to quickly attribute an LLM's text generation to training data points to inspect model behaviors, enhance its trustworthiness, and compare model-generated text with user-provided text. We describe the visual and interactive design of our tool and highlight usage scenarios for LLaMA2 models fine-tuned with two different datasets: online articles about recent disasters and finance-related question-answer pairs. Thanks to LLM Attributor's broad support for computational notebooks, users can easily integrate it into their workflow to interactively visualize attributions of their models. For easier access and extensibility, we open-source LLM Attributor at https://github.com/poloclub/ LLM-Attribution. The video demo is available at https://youtu.be/mIG2MDQKQxM.
Open AccessLLM Attributor: Interactive Visual Attribution for LLM Generation
While large language models (LLMs) have shown remarkable capability to generate convincing text across diverse domains, concerns around its potential risks have highlighted the importance of understanding the rationale behind text generation. We present LLM Attributor, a Python library that provides interactive visualizations for training data attribution of an LLM's text generation. Our library offers a new way to quickly attribute an LLM's text generation to training data points to inspect model behaviors, enhance its trustworthiness, and compare model-generated text with user-provided text. We describe the visual and interactive design of our tool and highlight usage scenarios for LLaMA2 models fine-tuned with two different datasets: online articles about recent disasters and finance-related question-answer pairs. Thanks to LLM Attributor's broad support for computational notebooks, users can easily integrate it into their workflow to interactively visualize attributions of their models. For easier access and extensibility, we open-source LLM Attributor at https://github.com/poloclub/ LLM-Attribution. The video demo is available at https://youtu.be/mIG2MDQKQxM.
- Open Access
Robust Principles: Architectural Design Principles for Adversarially Robust CNNs
Our research aims to unify existing works' diverging opinions on how architectural components affect the adversarial robustness of CNNs. To accomplish our goal, we synthesize a suite of three generalizable robust architectural design principles: (a) optimal range for depth and width configurations, (b) preferring convolutional over patchify stem stage, and (c) robust residual block design through adopting squeeze and excitation blocks and non-parametric smooth activation functions. Through extensive experiments across a wide spectrum of dataset scales, adversarial training methods, model parameters, and network design spaces, our principles consistently and markedly improve AutoAttack accuracy: 1-3 percentage points (pp) on CIFAR-10 and CIFAR-100, and 4-9 pp on ImageNet. The code is publicly available at https://github.com/poloclub/robust-principles.
Open AccessRobust Principles: Architectural Design Principles for Adversarially Robust CNNs
Our research aims to unify existing works' diverging opinions on how architectural components affect the adversarial robustness of CNNs. To accomplish our goal, we synthesize a suite of three generalizable robust architectural design principles: (a) optimal range for depth and width configurations, (b) preferring convolutional over patchify stem stage, and (c) robust residual block design through adopting squeeze and excitation blocks and non-parametric smooth activation functions. Through extensive experiments across a wide spectrum of dataset scales, adversarial training methods, model parameters, and network design spaces, our principles consistently and markedly improve AutoAttack accuracy: 1-3 percentage points (pp) on CIFAR-10 and CIFAR-100, and 4-9 pp on ImageNet. The code is publicly available at https://github.com/poloclub/robust-principles.