Brandon Dang

Nil-Jana Akpinar, Sandeep Avula, CJ Lee et al.

Large Language Models (LLMs) are increasingly used to generate and edit scientific abstracts, yet their integration into academic writing raises questions about trust, quality, and disclosure. Despite growing adoption, little is known about how readers perceive LLM-generated summaries and how these perceptions influence evaluations of scientific work. This paper presents a mixed-methods survey experiment investigating whether readers with ML expertise can distinguish between human- and LLM-generated abstracts, how actual and perceived LLM involvement affects judgments of quality and trustworthiness, and what orientations readers adopt toward AI-assisted writing. Our findings show that participants struggle to reliably identify LLM-generated content, yet their beliefs about LLM involvement significantly shape their evaluations. Notably, abstracts edited by LLMs are rated more favorably than those written solely by humans or LLMs. We also identify three distinct reader orientations toward LLM-assisted writing, offering insights into evolving norms and informing policy around disclosure and acceptable use in scientific communication.

Alicia Sagae, Chia-Jung Lee, Sandeep Avula et al.

Current methods for evaluating large language models (LLMs) typically focus on high-level tasks such as text generation, without targeting a particular AI application. This approach is not sufficient for evaluating LLMs for Responsible AI dimensions like fairness, since protected attributes that are highly relevant in one application may be less relevant in another. In this work, we construct a dataset that is driven by a real-world application (generate a plain-text product description, given a list of product features), parameterized by fairness attributes intersected with gendered adjectives and product categories, yielding a rich set of labeled prompts. We show how to use the data to identify quality, veracity, safety, and fairness gaps in LLMs, contributing a proposal for LLM evaluation paired with a concrete resource for the research community.

Brandon Dang, Martin J. Riedl, Matthew Lease

Though detection systems have been developed to identify obscene content such as pornography and violence, artificial intelligence is simply not good enough to fully automate this task yet. Due to the need for manual verification, social media companies may hire internal reviewers, contract specialized workers from third parties, or outsource to online labor markets for the purpose of commercial content moderation. These content moderators are often fully exposed to extreme content and may suffer lasting psychological and emotional damage. In this work, we aim to alleviate this problem by investigating the following question: How can we reveal the minimum amount of information to a human reviewer such that an objectionable image can still be correctly identified? We design and conduct experiments in which blurred graphic and non-graphic images are filtered by human moderators on Amazon Mechanical Turk (AMT). We observe how obfuscation affects the moderation experience with respect to image classification accuracy, interface usability, and worker emotional well-being.

Ye Zhang, Md Mustafizur Rahman, Alex Braylan et al.

A recent "third wave" of Neural Network (NN) approaches now delivers state-of-the-art performance in many machine learning tasks, spanning speech recognition, computer vision, and natural language processing. Because these modern NNs often comprise multiple interconnected layers, this new NN research is often referred to as deep learning. Stemming from this tide of NN work, a number of researchers have recently begun to investigate NN approaches to Information Retrieval (IR). While deep NNs have yet to achieve the same level of success in IR as seen in other areas, the recent surge of interest and work in NNs for IR suggest that this state of affairs may be quickly changing. In this work, we survey the current landscape of Neural IR research, paying special attention to the use of learned representations of queries and documents (i.e., neural embeddings). We highlight the successes of neural IR thus far, catalog obstacles to its wider adoption, and suggest potentially promising directions for future research.

Brandon Dang, Miles Hutson, Matt Lease

Internal HITs on Mechanical Turk can be programmatically restrictive, and as a result, many requesters turn to using external HITs as a more flexible alternative. However, creating such HITs can be redundant and time-consuming. We present MmmTurkey, a framework that enables researchers to not only quickly create and manage external HITs, but more significantly also capture and record detailed worker behavioral data characterizing how each worker completes a given task.