From the 1 of 5 linked papers with an AI index.
5 papers
Tag Questions and the Generational Reversal of Sycophancy Across 45 Language Models
Tapan Parikh
Appending a two-word confirmation tag to a decision question -- "Is X the better choice?" versus "X is the better choice, right?" -- changes whether a language model endorses the c…
Structured Output Collapses Answer Diversity Across 44 Language Models
Tapan Parikh
When a language model must choose one answer from a large space of equally valid options, a format clause -- "Reply with JSON only" -- changes which answer it chooses. We re-run th…
The One-Word Census: Answer-Choice Conformity Across 44 Language Models
Tapan Parikh
The paper introduces the One-Word Census, a set of 31 simple prompts used to measure how 44 language models converge on the same single-word answers, and analyzes the degree of ans…
The Communal Loom: Integrating Tangible Interaction and Participatory Data Collection for Assessing Well-Being
Niti Parikh, Yiran Zhao, Maria Alinea-Bravo +1
For most health or well-being interventions, the process of evaluation is distinct from the activity itself, both in terms of who is involved, and how the actual data is collected…
'Since Lawyers are Males..': Examining Implicit Gender Bias in Hindi Language Generation by LLMs
Ishika Joshi, Ishita Gupta, Adrita Dey +1
Large Language Models (LLMs) are increasingly being used to generate text across various languages, for tasks such as translation, customer support, and education. Despite these ad…