15 citations · 50 across the 24 of their papers we have counts for
28 papers
PoisonForge: Task-Level Targeted Poisoning Benchmark for Instruction-Tuned LLMs
Luze Sun, Anshuman Suri, Harsh Chaudhari +2
When practitioners fine-tune LLMs on unvetted datasets, an adversary can exploit the data supply chain through task-level poisoning: inserting a small number of crafted instruction…
20/20 Vision Language Models: A Prescription for Better VLMs through Data Curation Alone
DatologyAI, :, Siddharth Joshi +32
Data curation has shifted the quality-compute frontier for language-model and contrastive image-text pretraining, but its role for vision-language models (VLMs) is far less establi…
Toward a Principled Framework for Agent Safety Measurement
Shuyi Lin, Anshuman Suri, Alina Oprea +1
LLM agents emit actions, not just text, and once taken, those actions often cannot be undone. Yet today's agent-safety evaluations run greedy or a few sampled rollouts and report a…
The Finetuner's Fallacy: When to Pretrain with Your Finetuning Data
Christina Baek, Ricardo Pio Monti, David Schwab +31
Real-world model deployments demand strong performance on narrow domains where data is often scarce. Typically, practitioners finetune models to specialize them, but this risks ove…
ÜberWeb: Insights from Multilingual Curation for a 20-Trillion-Token Dataset
DatologyAI, :, Aldo Gael Carranza +32
Multilinguality is a core capability for modern foundation models, yet training high-quality multilingual models remains challenging due to uneven data availability across language…
Identifying Models Behind Text-to-Image Leaderboards
Ali Naseh, Yuefeng Peng, Anshuman Suri +3
Text-to-image (T2I) models are increasingly popular, producing a large share of AI-generated images online. To compare model quality, voting-based leaderboards have become the stan…
Text-to-Image Models Leave Identifiable Signatures: Implications for Leaderboard Security
Ali Naseh, Anshuman Suri, Yuefeng Peng +3
Generative AI leaderboards are central to evaluating model capabilities, but remain vulnerable to manipulation. Among key adversarial objectives is rank manipulation, where an atta…
Exploiting Leaderboards for Large-Scale Distribution of Malicious Models
Anshuman Suri, Harsh Chaudhari, Yuefeng Peng +3
While poisoning attacks on machine learning models have been extensively studied, the mechanisms by which adversaries can distribute poisoned models at scale remain largely unexplo…
Toward Principled LLM Safety Testing: Solving the Jailbreak Oracle Problem
Shuyi Lin, Anshuman Suri, Alina Oprea +1
As large language models (LLMs) become increasingly deployed in safety-critical applications, the lack of systematic methods to assess their vulnerability to jailbreak attacks pres…
SAGA: A Security Architecture for Governing AI Agentic Systems
Georgios Syros, Anshuman Suri, Jacob Ginesin +2
Large Language Model (LLM)-based agents increasingly interact, collaborate, and delegate tasks to one another autonomously with minimal human interaction. Industry guidelines for a…
DROP: Poison Dilution via Knowledge Distillation for Federated Learning
Georgios Syros, Anshuman Suri, Farinaz Koushanfar +2
Federated Learning is vulnerable to adversarial manipulation, where malicious clients can inject poisoned updates to influence the global model's behavior. While existing defense m…
Riddle Me This! Stealthy Membership Inference for Retrieval-Augmented Generation
Ali Naseh, Yuefeng Peng, Anshuman Suri +3
Retrieval-Augmented Generation (RAG) enables Large Language Models (LLMs) to generate grounded responses by leveraging external knowledge databases without altering model parameter…
Do Parameters Reveal More than Loss for Membership Inference?
Anshuman Suri, Xiao Zhang, David Evans
Membership inference attacks are used as a key tool for disclosure auditing. They aim to infer whether an individual record was used to train a model. While such evaluations are us…
Do Membership Inference Attacks Work on Large Language Models?
Michael Duan, Anshuman Suri, Niloofar Mireshghallah +7
Membership inference attacks (MIAs) attempt to predict whether a particular datapoint is a member of a target model's training data. Despite extensive research on traditional machi…
SoK: Memorization in General-Purpose Large Language Models
Valentin Hartmann, Anshuman Suri, Vincent Bindschaedler +3
Large Language Models (LLMs) are advancing at a remarkable pace, with myriad applications under development. Unlike most earlier machine learning models, they are no longer built f…
SoK: Pitfalls in Evaluating Black-Box Attacks
Fnu Suya, Anshuman Suri, Tingwei Zhang +3
Numerous works study black-box attacks on image classifiers. However, these works make different assumptions on the adversary's knowledge and current literature lacks a cohesive or…
Manipulating Transfer Learning for Property Inference
Yulong Tian, Fnu Suya, Anshuman Suri +2
Transfer learning is a popular method for tuning pretrained (upstream) models for different downstream tasks using limited data and computational resources. We study how an adversa…
SoK: Let the Privacy Games Begin! A Unified Treatment of Data Inference Privacy in Machine Learning
Ahmed Salem, Giovanni Cherubin, David Evans +5
Deploying machine learning models in production may allow adversaries to infer sensitive information about training data. There is a vast literature analyzing different types of in…
Dissecting Distribution Inference
Anshuman Suri, Yifu Lu, Yanjin Chen +1
A distribution inference attack aims to infer statistical properties of data used to train machine learning models. These attacks are sometimes surprisingly potent, but the factors…
Subject Membership Inference Attacks in Federated Learning
Anshuman Suri, Pallika Kanani, Virendra J. Marathe +1
Privacy attacks on Machine Learning (ML) models often focus on inferring the existence of particular data points in the training data. However, what the adversary really wants to k…
Formalizing and Estimating Distribution Inference Risks
Anshuman Suri, David Evans
Distribution inference, sometimes called property inference, infers statistical properties about a training set from access to a model trained on that data. Distribution inference…
Formalizing Distribution Inference Risks
Anshuman Suri, David Evans
Property inference attacks reveal statistical properties about a training set but are difficult to distinguish from the primary purposes of statistical machine learning, which is t…
QnAMaker: Data to Bot in 2 Minutes
Parag Agrawal, Tulasi Menon, Aya Kamel +15
Having a bot for seamless conversations is a much-desired feature that products and services today seek for their websites and mobile apps. These bots help reduce traffic received…
One Neuron to Fool Them All
Anshuman Suri, David Evans
Despite vast research in adversarial examples, the root causes of model susceptibility are not well understood. Instead of looking at attack-specific robustness, we propose a notio…