Search Results for author:
Found 19 papers, 12 papers with code
A False Sense of Privacy: Evaluating Textual Data Sanitization Beyond Surface-level Privacy Leakage
Sanitizing sensitive text data typically involves removing personally identifiable information (PII) or generating synthetic data under the assumption that these methods adequately protect privacy; however, their effectiveness is often only assessed by measuring the leakage of explicit identifiers but ignoring nuanced textual markers that can lead to re-identification.
ParaPO: Aligning Language Models to Reduce Verbatim Reproduction of Pre-training Data
When applied to the instruction-tuned Tulu3-8B model, ParaPO with system prompting successfully preserves famous quotation recall while reducing unintentional regurgitation (from 8. 7 to 6. 3 in creative writing) when prompted not to regurgitate.
Privacy Ripple Effects from Adding or Removing Personal Information in Language Model Training
Due to the sensitive nature of personally identifiable information (PII), its owners may have the authority to control its inclusion or request its removal from large-language model (LLM) training.
Synthetic Data Can Mislead Evaluations: Membership Inference as Machine Text Detection
Recent work shows membership inference attacks (MIAs) on large language models (LLMs) produce inconclusive results, partly due to difficulties in creating non-member datasets without temporal shifts.
Machine Unlearning Doesn't Do What You Think: Lessons for Generative AI Policy, Research, and Practice
We articulate fundamental mismatches between technical methods for machine unlearning in Generative AI, and documented aspirations for broader impact that these methods could have for law and policy.
Differentially Private Learning Needs Better Model Initialization and Self-Distillation
Differentially private SGD (DPSGD) enables privacy-preserving training of language models, but often reduces utility, diversity, and linguistic quality.
AI as Humanity's Salieri: Quantifying Linguistic Creativity of Language Models via Systematic Attribution of Machine Text against Web Text
We present CREATIVITY INDEX as the first step to quantify the linguistic creativity of a text by reconstructing it from existing text snippets on the web.
HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Human-AI Interactions
AI agents are increasingly autonomous in their interactions with human users and tools, leading to increased interactional safety risks.
Trust No Bot: Discovering Personal Disclosures in Human-LLM Conversations in the Wild
Measuring personal disclosures made in human-chatbot interactions can provide a better understanding of users' AI literacy and facilitate privacy research for large language models (LLMs).
CopyBench: Measuring Literal and Non-Literal Reproduction of Copyright-Protected Text in Language Model Generation
Evaluating the degree of reproduction of copyright-protected content by language models (LMs) is of significant interest to the AI and legal communities.
WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models
As WildJailbreak considerably upgrades the quality and scale of existing safety resources, it uniquely enables us to examine the scaling effects of data and the interplay of data properties and model capabilities during safety training.
Developing Story: Case Studies of Generative AI's Use in Journalism
Journalists are among the many users of large language models (LLMs).
Alpaca against Vicuna: Using LLMs to Uncover Memorization of LLMs
In this paper, we introduce a black-box prompt optimization method that uses an attacker LLM agent to uncover higher levels of memorization in a victim agent, compared to what is revealed by prompting the target model with the training data directly, which is the dominant approach of quantifying memorization in LLMs.
Do Membership Inference Attacks Work on Large Language Models?
Membership inference attacks (MIAs) attempt to predict whether a particular datapoint is a member of a target model's training data.
A Roadmap to Pluralistic Alignment
We identify and formalize three possible ways to define and operationalize pluralism in AI systems: 1) Overton pluralistic models that present a spectrum of reasonable responses; 2) Steerably pluralistic models that can steer to reflect certain perspectives; and 3) Distributionally pluralistic models that are well-calibrated to a given population in distribution.
A Block Metropolis-Hastings Sampler for Controllable Energy-based Text Generation
Recent work has shown that energy-based language modeling is an effective framework for controllable text generation because it enables flexible integration of arbitrary discriminators.
Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theory
The interactive use of large language models (LLMs) in AI assistants (at work, home, etc.)
Misusing Tools in Large Language Models With Visual Adversarial Examples
Large Language Models (LLMs) are being enhanced with the ability to use tools and to process multiple modalities.
Smaller Language Models are Better Black-box Machine-Generated Text Detectors
With the advent of fluent generative language models that can produce convincing utterances very similar to those written by humans, distinguishing whether a piece of text is machine-generated or human-written becomes more challenging and more important, as such models could be used to spread misinformation, fake news, fake reviews and to mimic certain authors and figures.
