A neural mechanism for conserved value computations integrating information and rewards

Area of Science:

• Neuroscience
• Decision Science
• Behavioral Economics

Background:

• Decision-making theories typically focus on choices with objective values.
• Humans and animals often seek information with uncertain future rewards, suggesting information itself has subjective value.
• The neural basis for integrating information value with extrinsic rewards remains largely unknown.

Purpose of the Study:

• To investigate the conserved computational principles underlying value judgments in multi-attribute decisions.
• To identify the neural substrates responsible for integrating information value and extrinsic rewards.
• To elucidate the mechanisms by which the brain decides to seek information about the future.

Main Methods:

• Comparative analysis of value judgments in humans and monkeys during information-seeking tasks.
• Electrophysiological recordings in the lateral habenula (LHb) and its input areas.
• Computational modeling of decision-making processes.

Main Results:

• Human and monkey value judgments exhibit conserved computational principles when trading off information and extrinsic rewards.
• Lateral habenula (LHb) neurons signal integrated subjective value, combining information and extrinsic reward values.
• The LHb causally influences decisions to seek information, predicting ongoing choices.
• Input areas to the LHb primarily signal components, not integrated value.

Conclusions:

• The lateral habenula (LHb) is a key neural substrate for integrating information value with extrinsic rewards.
• Conserved neural mechanisms in the LHb underlie the computation of subjective value for information seeking.
• This research reveals how the brain makes decisions to acquire knowledge about future outcomes.