Explanipedia

Machine Text Detectors are Membership Inference Attacks Open

Ryuto Koike, Liam Dugan, Masahiro Kaneko, Naoaki Okazaki · 2025

Although membership inference attacks (MIAs) and machine-generated text detection target different goals, identifying training samples and synthetic texts, their methods often exploit similar signals based on a language model's probability…

Controlling Difficulty of Generated Text for AI-Assisted Language Learning Open

Min Jin, Liam Dugan, Chris Callison-Burch · 2025

Practicing conversations with large language models (LLMs) presents a promising alternative to traditional in-person language learning. However, most LLMs generate text at a near-native level of complexity, making them ill-suited for begin…

Domain Gating Ensemble Networks for AI-Generated Text Detection Open

A.K. Tripathi, Liam Dugan, Charis Gao, Maggie Huan, Eekseong Jin , et al. · 2025

As state-of-the-art language models continue to improve, the need for robust detection of machine-generated text becomes increasingly critical. However, current state-of-the-art machine text detectors struggle to adapt to new unseen domain…

Group-Adaptive Threshold Optimization for Robust AI-Generated Text Detection Open

Martin Jung, Cynthia Fuertes Panizo, Liam Dugan, May Fung, Pin‐Yu Chen , et al. · 2025

The advancement of large language models (LLMs) has made it difficult to differentiate human-written text from AI-generated text. Several AI-text detectors have been developed in response, which typically utilize a fixed global threshold (…

GenAI Content Detection Task 3: Cross-Domain Machine-Generated Text Detection Challenge Open

Liam Dugan, Andrew Zhu, Firoj Alam, Preslav Nakov, Marianna Apidianaki , et al. · 2025

Recently there have been many shared tasks targeting the detection of generated text from Large Language Models (LLMs). However, these shared tasks tend to focus either on cases where text is limited to one particular domain or cases where…

MiRAGeNews: Multimodal Realistic AI-Generated News Detection Open

Rongqin Huang, Liam Dugan, Yue Yang, Chris Callison-Burch · 2024

The proliferation of inflammatory or misleading "fake" news content has become increasingly common in recent years. Simultaneously, it has become easier than ever to use AI tools to generate photorealistic images depicting any scene imagin…

ReDel: A Toolkit for LLM-Powered Recursive Multi-Agent Systems Open

Andrew Zhu, Liam Dugan, Chris Callison-Burch · 2024

Recently, there has been increasing interest in using Large Language Models (LLMs) to construct complex multi-agent systems to perform tasks such as compiling literature reviews, drafting consumer reports, and planning vacations. Many tool…

RAID: A Shared Benchmark for Robust Evaluation of Machine-Generated Text Detectors Open

Liam Dugan, Alyssa Hwang, Filip Trhlík, Josh Magnus Ludan, Andrew Zhu , et al. · 2024

Many commercial and open-source models claim to detect machine-generated text with extremely high accuracy (99% or more). However, very few of these detectors are evaluated on shared benchmark datasets and even when they are, the datasets …

FanOutQA: A Multi-Hop, Multi-Document Question Answering Benchmark for Large Language Models Open

Andrew Zhu, Alyssa Hwang, Liam Dugan, Chris Callison-Burch · 2024

One type of question that is commonly found in day-to-day scenarios is ``fan-out'' questions, complex multi-hop, multi-document reasoning questions that require finding information about a large number of entities. However, there exist few…

Interpretable-by-Design Text Understanding with Iteratively Generated Concept Bottleneck Open

Josh Magnus Ludan, Qing Lyu, Yue Yang, Liam Dugan, Mark Yatskar , et al. · 2023

Black-box deep neural networks excel in text classification, yet their application in high-stakes domains is hindered by their lack of interpretability. To address this, we propose Text Bottleneck Models (TBM), an intrinsically interpretab…

Kani: A Lightweight and Highly Hackable Framework for Building Language Model Applications Open

Andrew Zhu, Liam Dugan, Alyssa Hwang, Chris Callison-Burch · 2023

Language model applications are becoming increasingly popular and complex, often including features like tool usage and retrieval augmentation. However, existing frameworks for such applications are often opinionated, deciding for develope…

Real or Fake Text?: Investigating Human Ability to Detect Boundaries between Human-Written and Machine-Generated Text Open

Liam Dugan, Daphne Ippolito, Arun Kirubarajan, Sherry Shi, Chris Callison-Burch · 2023

As text generated by large language models proliferates, it becomes vital to understand how humans engage with such text, and whether or not they are able to detect when the text they are reading did not originate with a human writer. Prio…

Learning When to Speak: Latency and Quality Trade-offs for Simultaneous Speech-to-Speech Translation with Offline Models Open

Liam Dugan, Anshul Wadhawan, Kyle Spence, Chris Callison-Burch, Morgan McGuire , et al. · 2023

Recent work in speech-to-speech translation (S2ST) has focused primarily on offline settings, where the full input utterance is available before any output is given. This, however, is not reasonable in many real-world scenarios. In latency…

Exploring the Curious Case of Code Prompts Open

Li Zhang, Liam Dugan, Hainiu Xu, Chris Callison-Burch · 2023

Recent work has shown that prompting language models with code-like representations of natural language leads to performance improvements on structured reasoning tasks. However, such tasks comprise only a small subset of all natural langua…

Exploring the Curious Case of Code Prompts Open

Li Zhang, Liam Dugan, Hainiu Xu, Chris Callison-Burch · 2023

Recent work has shown that prompting language models with code-like representations of natural language leads to performance improvements on structured reasoning tasks. However, such tasks comprise only a small subset of all natural langua…

Kani: A Lightweight and Highly Hackable Framework for Building Language Model Applications Open

Andrew Zhu, Liam Dugan, Alyssa Hwang, Chris Callison-Burch · 2023

Language model applications are becoming increasingly popular and complex, often including features like tool usage and retrieval augmentation. However, existing frameworks for such applications are often opinionated, deciding for develope…

Enhancing Human Summaries for Question-Answer Generation in Education Open

Hannah Gonzalez, Liam Dugan, Eleni Miltsakaki, Zhiqi Cui, Jiaxuan Ren , et al. · 2023

Hannah Gonzalez, Liam Dugan, Eleni Miltsakaki, Zhiqi Cui, Jiaxuan Ren, Bryan Li, Shriyash Upadhyay, Etan Ginsberg, Chris Callison-Burch. Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2…

Real or Fake Text?: Investigating Human Ability to Detect Boundaries Between Human-Written and Machine-Generated Text Open

Liam Dugan, Daphne Ippolito, Arun Kirubarajan, Sherry Shi, Chris Callison-Burch · 2022

As text generated by large language models proliferates, it becomes vital to understand how humans engage with such text, and whether or not they are able to detect when the text they are reading did not originate with a human writer. Prio…

The Case for a Single Model that can Both Generate Continuations and Fill in the Blank Open

Daphne Ippolito, Liam Dugan, Emily Reif, Ann Yuan, Andy Coenen , et al. · 2022

The task of inserting text into a specified position in a passage, known as fill in the blank (FitB), is useful for a variety of applications where writers interact with a natural language generation (NLG) system to craft text. While previ…

A Feasibility Study of Answer-Agnostic Question Generation for Education Open

Liam Dugan, Eleni Miltsakaki, Shriyash Upadhyay, Etan Ginsberg, Hannah Gonzalez , et al. · 2022

We conduct a feasibility study into the applicability of answer-agnostic question generation models to textbook passages. We show that a significant portion of errors in such systems arise from asking irrelevant or uninterpretable question…

The Case for a Single Model that can Both Generate Continuations and Fill-in-the-Blank Open

Daphne Ippolito, Liam Dugan, Emily Reif, Ann Yuan, Andy Coenen , et al. · 2022

The task of inserting text into a specified position in a passage, known as fill in the blank (FitB), is useful for a variety of applications where writers interact with a natural language generation (NLG) system to craft text. While previ…

A Feasibility Study of Answer-Unaware Question Generation for Education Open

Liam Dugan, Eleni Miltsakaki, Shriyash Upadhyay, Etan Ginsberg, Hannah Gonzalez , et al. · 2022

Liam Dugan, Eleni Miltsakaki, Shriyash Upadhyay, Etan Ginsberg, Hannah Gonzalez, DaHyeon Choi, Chuning Yuan, Chris Callison-Burch. Findings of the Association for Computational Linguistics: ACL 2022. 2022.

Watching Paint Dry: I/VOC Emissions from Architectural Coatings and their Impact on SOA Formation Open

Rebecca Tanzer-Gruener, Liam Dugan, Mark E. Bier, Allen Robinson, Albert A. Presto · 2021

Emissions from volatile chemical products (VCPs) are emerging as a major source of anthropogenic secondary organic aerosol (SOA) precursors. Paints and coatings are an important class of VCPs that emit both volatile and intermediate volati…

Watching Paint Dry: I/VOC Emissions from Architectural Coatings and their Impact on SOA Formation Open

Rebecca Tanzer-Gruener, Liam Dugan, Mark E. Bier, Allen L. Robinson, Albert A. Presto · 2021

Emissions from volatile chemical products (VCPs) are emerging as a major source of anthropogenic secondary organic aerosol (SOA) precursors. Paints and coatings are an important class of VCPs that emit both volatile and intermediate volati…

RoFT: A Tool for Evaluating Human Detection of Machine-Generated Text Open

Liam Dugan, Daphne Ippolito, Arun Kirubarajan, Chris Callison-Burch · 2020

In recent years, large neural networks for natural language generation (NLG) have made leaps and bounds in their ability to generate fluent text. However, the tasks of evaluating quality differences between NLG systems and understanding ho…

Liam Dugan YOU? Author Swipe