analiza aiqbench
This research introduces "epistemic diversity" as a critical metric for evaluating large language models, arguing that accuracy alone is insufficient when systems support inquiry and teaching. The study reveals that frontier models often exhibit "narrowness," collapsing vast spaces of valid reasoning into limited canonical answers, which can restrict user access to alternative knowledge paths. Consequently, the field must expand its evaluation frameworks beyond simple correctness to include the breadth of explanations and reasoning routes presented by AI systems.
On Epistemic Diversity in Large Language Models
arXiv:2609.04835v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used not only to retrieve information, but to answer questions, explain, teach, and support inquiry. In such settings, evaluation cannot be exhausted by accuracy or alignment alone. A system may give a correct answer while still narrowing users' access %to knowledge. to alternative valid answers, explanations, or reasoning routes. Drawing on the broader notion of epistemic diversity in philosophy and social epistemology, we formalize it in the context of LLMs as the range of valid answers, explanations, and reasoning routes that an LLM exposes to users. We argue that epistemic diversity is a useful evaluation dimension for settings where LLMs are used to support knowledge-intensive tasks. We propose a preliminary framework for conceptualizing and measuring epistemic diversity in LLMs, and operationalize it in two domains. We find that frontier LLMs often exhibit epistemic narrowness, repeatedly collapsing large valid answer spaces onto small canonical subsets. These findings suggest that LLM evaluation should move beyond accuracy-oriented paradigms and treat epistemic diversity as an important dimension of model capability.
Abstract:Large language models (LLMs) are increasingly used not only to retrieve information, but to answer questions, explain, teach, and support inquiry. In such settings, evaluation cannot be exhausted by accuracy or alignment alone. A system may give a correct answer while still narrowing users' access %to knowledge. to alternative valid answers, explanations, or reasoning routes. Drawing on the broader notion of epistemic diversity in philosophy and social epistemology, we formalize it in the context of LLMs as the range of valid answers, explanations, and reasoning routes that an LLM exposes to users. We argue that epistemic diversity is a useful evaluation dimension for settings where LLMs are used to support knowledge-intensive tasks. We propose a preliminary framework for conceptualizing and measuring epistemic diversity in LLMs, and operationalize it in two domains. We find that frontier LLMs often exhibit epistemic narrowness, repeatedly collapsing large valid answer spaces onto small canonical subsets. These findings suggest that LLM evaluation should move beyond accuracy-oriented paradigms and treat epistemic diversity as an important dimension of model capability.
Computation and Language (cs.CL)
arXiv:2609.04835 [cs.CL]
(or arXiv:2609.04835v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2609.04835 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 267 cuvinte
Subiecte conexe
The Continual Constitutional Convention: How language models revise the documents that govern them
AI Alignment Forum · 13 oct. 2026
The AI Epistemic Deference Index: A Continuous Measure of Sycophancy
arXiv · 9 iun. 2026
Do Large Language Models Hallucinate Electric Fata Morganas?
arXiv · 21 aug. 2026
Prompting the Unknown: Understanding Response Uncertainty in Large Language Models
arXiv · 3 sept. 2026