Reward Valuation in Large Language Models: Causal Induction of Anhedonia
Abstract
Recent frontier models mimic complex aspects of human cognition. Here we ask whether this alignment extends into reward valuation, which we assess in a mechanistic framework. Specifically, we use clinical tests that were developed to evaluate anhedonia in human subjects with major depressive disorders. Mechanistically, anhedonia is frequently associated with dysregulation in the Nucleus Accumbens (NAc) and the broader dopaminergic reward system. While neuroimaging has localized these deficits, establishing a causal link between NAc activity and specific behavioral symptoms remains a challenge. We use these ideas from neuroscience to functionally identify reward-anticipatory units in state-of-the-art AI models, and evaluate their causal involvement via targeted perturbations. We find that not only are such model units predictive of NAc brain recordings, their perturbation also induces behavioral effects mirroring human anhedonia: the model opts for low-effort, low-reward tasks in effort-based decision-making paradigms. Crucially, our results demonstrate that this represents a specific deficit in self-centered reward valuation and anticipation--rather than a loss of task capability, reward calculation, or effort avoidance. This induced vulnerability aligns with clinical measures of anhedonia and motivation in humans, such as DARS and MAP-SR, instruments that contain no reward-related vocabulary, ruling out a purely lexical account of the perturbation effect. Taken together, our results suggest reward valuation circuits in AI models that functionally mimic those in humans.
Related
- Vision-language representations link food perception to subjective valuation
- Causal event structure shapes convergent reactivation signatures in human cortex and language model
- Towards Sparse Causal Features for Zero-shot Mutation Effect Prediction in a Protein Language Model
- The Use of Causal Language in Case Reports and Case Series Reporting Ischemic Stroke Following Cervical Manual Therapy: A Qualitative Study Protocol
- Expertise modifies neural responses to causal inference during language comprehension
BibTeX
@misc{honarmand2026reward,
title = {{Reward Valuation in Large Language Models: Causal Induction of Anhedonia}},
author = {Melika Honarmand and Samin Mahdipour Aghabagher and Martin Schrimpf},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2607.06626},
eprint = {2607.06626},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2607.06626},
note = {CC BY-NC-SA 4.0},
}
RIS
TY - UNPB
TI - Reward Valuation in Large Language Models: Causal Induction of Anhedonia
AU - Melika Honarmand
AU - Samin Mahdipour Aghabagher
AU - Martin Schrimpf
PY - 2026
DA - 2026/09/30
PB - arXiv
DO - 10.48550/arxiv.2607.06626
UR - https://arxiv.org/abs/2607.06626
AB - Recent frontier models mimic complex aspects of human cognition. Here we ask whether this alignment extends into reward valuation, which we assess in a mechanistic framework. Specifically, we use clinical tests that were developed to evaluate anhedonia in human subjects with major depressive disorders. Mechanistically, anhedonia is frequently associated with dysregulation in the Nucleus Accumbens (NAc) and the broader dopaminergic reward system. While neuroimaging has localized these deficits, establishing a causal link between NAc activity and specific behavioral symptoms remains a challenge. We use these ideas from neuroscience to functionally identify reward-anticipatory units in state-of-the-art AI models, and evaluate their causal involvement via targeted perturbations. We find that not only are such model units predictive of NAc brain recordings, their perturbation also induces behavioral effects mirroring human anhedonia: the model opts for low-effort, low-reward tasks in effort-based decision-making paradigms. Crucially, our results demonstrate that this represents a specific deficit in self-centered reward valuation and anticipation--rather than a loss of task capability, reward calculation, or effort avoidance. This induced vulnerability aligns with clinical measures of anhedonia and motivation in humans, such as DARS and MAP-SR, instruments that contain no reward-related vocabulary, ruling out a purely lexical account of the perturbation effect. Taken together, our results suggest reward valuation circuits in AI models that functionally mimic those in humans.
ER -