5 papers
Alignment has a Fantasia Problem
Nathanael Jo, Zoe De Simone, Mitchell Gordon +1
In accomplishing complex tasks, human cognition typically progresses from abstract to concrete (e.g., from brainstorming ideas to writing an essay). With the advent of highly capab…
Task-Dependent Evaluation of LLM Output Homogenization: A Taxonomy-Guided Framework
Shomik Jain, Jack Lanchantin, Maximilian Nickel +4
Large language models often generate homogeneous outputs, but whether this is problematic depends on the specific task. For objective math tasks, responses may vary in terms of pro…
Interaction Context Often Increases Sycophancy in LLMs
Shomik Jain, Charlotte Park, Matt Viana +2
We investigate how the presence and type of interaction context shapes sycophancy in LLMs. While real-world interactions allow models to mirror a user's values, preferences, and se…
Allocation Multiplicity: Evaluating the Promises of the Rashomon Set
Shomik Jain, Margaret Wang, Kathleen Creel +1
The Rashomon set of equally-good models promises less discriminatory algorithms, reduced outcome homogenization, and fairer decisions through model ensembles or reconciliation. How…
Homogeneous Algorithms Can Reduce Competition in Personalized Pricing
Nathanael Jo, Kathleen Creel, Ashia Wilson +1
Firms' algorithm development practices are often homogeneous. Whether firms train algorithms on similar data, aim at similar benchmarks, or rely on similar pre-trained models, the…