-
Anthropomorphism and Trust in Human-Large Language Model interactions
Authors:
Akila Kadambi,
Ylenia D'Elia,
Tanishka Shah,
Iulia Comsa,
Alison Lentz,
Katie Siri-Ngammuang,
Tara Buechler,
Jonas Kaplan,
Antonio Damasio,
Srini Narayanan,
Lisa Aziz-Zadeh
Abstract:
With large language models (LLMs) becoming increasingly prevalent in daily life, so too has the tendency to attribute to them human-like minds and emotions, or anthropomorphize them. Here, we investigate dimensions people use to anthropomorphize and attribute trust toward LLMs across more than 2,000 human-LLM interactions. Participants (N=115) engaged with LLM chatbots systematically varied in war…
▽ More
With large language models (LLMs) becoming increasingly prevalent in daily life, so too has the tendency to attribute to them human-like minds and emotions, or anthropomorphize them. Here, we investigate dimensions people use to anthropomorphize and attribute trust toward LLMs across more than 2,000 human-LLM interactions. Participants (N=115) engaged with LLM chatbots systematically varied in warmth (friendliness), competence (capability, coherence), and empathy (cognitive and affective). Warmth and cognitive empathy significantly predicted perceptions on all outcomes (perceived anthropomorphism, trust, similarity, relational closeness, frustration, usefulness), while competence predicted all outcomes except for anthropomorphism. Affective empathy primarily predicted perceived relational measures, but did not predict the epistemic outcomes. Topic sub-analyses showed that more subjective, personally relevant topics (e.g., relationship advice) amplified these effects, producing greater human-likeness and relational connection with the LLM than did objective topics. Together, these findings reveal that warmth, competence, and empathy are key dimensions through which people attribute relational and epistemic perceptions to artificial agents.
△ Less
Submitted 1 March, 2026;
originally announced April 2026.
-
A smoothed proximal trust-region algorithm for nonconvex optimization problems with $L^p$-regularization, $p\in (0,1)$
Authors:
Harbir Antil,
Anna Lentz
Abstract:
We investigate a trust-region algorithm to solve a nonconvex optimization problem with $L^p$-regularization for $p\in(0,1)$. The algorithm relies on descent properties of a so-called generalized Cauchy point that can be obtained efficiently by a line search along a suitable proximal path. To handle the nonconvexity and nonsmoothness of the $L^p$-pseudonorm, we replace it by a smooth approximation…
▽ More
We investigate a trust-region algorithm to solve a nonconvex optimization problem with $L^p$-regularization for $p\in(0,1)$. The algorithm relies on descent properties of a so-called generalized Cauchy point that can be obtained efficiently by a line search along a suitable proximal path. To handle the nonconvexity and nonsmoothness of the $L^p$-pseudonorm, we replace it by a smooth approximation and construct a convex upper bound of that approximation. This enables us to use results of a trust-region method for composite problems with a convex nonsmooth term. We prove convergence properties of the resulting smoothed proximal trust-region algorithm and investigate its performance in some numerical examples. Furthermore, approximate subproblem solvers for the arising trust-region subproblems are considered.
△ Less
Submitted 21 August, 2025;
originally announced August 2025.
-
Capacitary measures in fractional order Sobolev spaces: Compactness and applications to minimization problems
Authors:
Anna Lentz
Abstract:
Capacitary measures form a class of measures that vanish on sets of capacity zero. These measures are compact with respect to so-called $γ$-convergence, which relates a sequence of measures to the sequence of solutions of relaxed Dirichlet problems. This compactness result is already known for the classical $H^1(Ω)$-capacity. This paper extends it to the fractional capacity defined for fractional…
▽ More
Capacitary measures form a class of measures that vanish on sets of capacity zero. These measures are compact with respect to so-called $γ$-convergence, which relates a sequence of measures to the sequence of solutions of relaxed Dirichlet problems. This compactness result is already known for the classical $H^1(Ω)$-capacity. This paper extends it to the fractional capacity defined for fractional order Sobolev spaces $H^s(Ω)$ for $s\in (0,1)$. The compactness result is applied to obtain a finer optimality condition for a class of minimization problems in $H^s(Ω)$.
△ Less
Submitted 16 December, 2024;
originally announced December 2024.
-
LLMs achieve adult human performance on higher-order theory of mind tasks
Authors:
Winnie Street,
John Oliver Siy,
Geoff Keeling,
Adrien Baranes,
Benjamin Barnett,
Michael McKibben,
Tatenda Kanyere,
Alison Lentz,
Blaise Aguera y Arcas,
Robin I. M. Dunbar
Abstract:
This paper examines the extent to which large language models (LLMs) have developed higher-order theory of mind (ToM); the human ability to reason about multiple mental and emotional states in a recursive manner (e.g. I think that you believe that she knows). This paper builds on prior work by introducing a handwritten test suite -- Multi-Order Theory of Mind Q&A -- and using it to compare the per…
▽ More
This paper examines the extent to which large language models (LLMs) have developed higher-order theory of mind (ToM); the human ability to reason about multiple mental and emotional states in a recursive manner (e.g. I think that you believe that she knows). This paper builds on prior work by introducing a handwritten test suite -- Multi-Order Theory of Mind Q&A -- and using it to compare the performance of five LLMs to a newly gathered adult human benchmark. We find that GPT-4 and Flan-PaLM reach adult-level and near adult-level performance on ToM tasks overall, and that GPT-4 exceeds adult performance on 6th order inferences. Our results suggest that there is an interplay between model size and finetuning for the realisation of ToM abilities, and that the best-performing LLMs have developed a generalised capacity for ToM. Given the role that higher-order ToM plays in a wide range of cooperative and competitive human behaviours, these findings have significant implications for user-facing LLM applications.
△ Less
Submitted 31 May, 2024; v1 submitted 29 May, 2024;
originally announced May 2024.
-
The Ethics of Advanced AI Assistants
Authors:
Iason Gabriel,
Arianna Manzini,
Geoff Keeling,
Lisa Anne Hendricks,
Verena Rieser,
Hasan Iqbal,
Nenad Tomašev,
Ira Ktena,
Zachary Kenton,
Mikel Rodriguez,
Seliem El-Sayed,
Sasha Brown,
Canfer Akbulut,
Andrew Trask,
Edward Hughes,
A. Stevie Bergman,
Renee Shelby,
Nahema Marchal,
Conor Griffin,
Juan Mateos-Garcia,
Laura Weidinger,
Winnie Street,
Benjamin Lange,
Alex Ingerman,
Alison Lentz
, et al. (32 additional authors not shown)
Abstract:
This paper focuses on the opportunities and the ethical and societal risks posed by advanced AI assistants. We define advanced AI assistants as artificial agents with natural language interfaces, whose function is to plan and execute sequences of actions on behalf of a user, across one or more domains, in line with the user's expectations. The paper starts by considering the technology itself, pro…
▽ More
This paper focuses on the opportunities and the ethical and societal risks posed by advanced AI assistants. We define advanced AI assistants as artificial agents with natural language interfaces, whose function is to plan and execute sequences of actions on behalf of a user, across one or more domains, in line with the user's expectations. The paper starts by considering the technology itself, providing an overview of AI assistants, their technical foundations and potential range of applications. It then explores questions around AI value alignment, well-being, safety and malicious uses. Extending the circle of inquiry further, we next consider the relationship between advanced AI assistants and individual users in more detail, exploring topics such as manipulation and persuasion, anthropomorphism, appropriate relationships, trust and privacy. With this analysis in place, we consider the deployment of advanced assistants at a societal scale, focusing on cooperation, equity and access, misinformation, economic impact, the environment and how best to evaluate advanced AI assistants. Finally, we conclude by providing a range of recommendations for researchers, developers, policymakers and public stakeholders.
△ Less
Submitted 28 April, 2024; v1 submitted 24 April, 2024;
originally announced April 2024.
-
Spatially sparse optimization problems in fractional order Sobolev spaces
Authors:
Anna Lentz,
Daniel Wachsmuth
Abstract:
We investigate time-dependent optimization problems in fractional Sobolev spaces with the sparsity promoting $L^p$-pseudo norm for $0<p<1$ in the objective functional. In order to avoid computing the fractional Laplacian on the time-space cylinder $I\times Ω$, we introduce an auxiliary function $w$ on $Ω$ that is an upper bound for the function $u\in L^2(I\timesΩ)$. We prove existence and regulari…
▽ More
We investigate time-dependent optimization problems in fractional Sobolev spaces with the sparsity promoting $L^p$-pseudo norm for $0<p<1$ in the objective functional. In order to avoid computing the fractional Laplacian on the time-space cylinder $I\times Ω$, we introduce an auxiliary function $w$ on $Ω$ that is an upper bound for the function $u\in L^2(I\timesΩ)$. We prove existence and regularity results and derive a necessary optimality condition. This is done by smoothing the $L^p$-pseudo norm and by penalizing the inequality constraint regarding $u$ and $w$. The problem is solved numerically with an iterative scheme whose weak limit points satisfy a weaker form of the necessary optimality condition.
△ Less
Submitted 21 May, 2025; v1 submitted 22 February, 2024;
originally announced February 2024.
-
Engaging Engineering Teams Through Moral Imagination: A Bottom-Up Approach for Responsible Innovation and Ethical Culture Change in Technology Companies
Authors:
Benjamin Lange,
Geoff Keeling,
Amanda McCroskery,
Ben Zevenbergen,
Sandra Blascovich,
Kyle Pedersen,
Alison Lentz,
Blaise Aguera y Arcas
Abstract:
We propose a "Moral Imagination" methodology to facilitate a culture of responsible innovation for engineering and product teams in technology companies. Our approach has been operationalized over the past two years at Google, where we have conducted over 50 workshops with teams across the organization. We argue that our approach is a crucial complement to existing formal and informal initiatives…
▽ More
We propose a "Moral Imagination" methodology to facilitate a culture of responsible innovation for engineering and product teams in technology companies. Our approach has been operationalized over the past two years at Google, where we have conducted over 50 workshops with teams across the organization. We argue that our approach is a crucial complement to existing formal and informal initiatives for fostering a culture of ethical awareness, deliberation, and decision-making in technology design such as company principles, ethics and privacy review procedures, and compliance controls. We characterize some of the distinctive benefits of our methodology for the technology sector in particular.
△ Less
Submitted 28 October, 2023; v1 submitted 12 June, 2023;
originally announced June 2023.