Why This Matters:
- Why this matters: Reveals neural mechanisms that explain how people balance exploration and exploitation when facing potential rewards or losses.
- How this aligns with previous research: Extends prior work linking the amygdala to exploration and mood regulation, showing these processes also apply in loss-avoidance settings.
- Future implications: Could guide development of treatments for psychiatric conditions—such as depression and anxiety—where exploration strategies are disrupted.
Summary: A recent study using single-neuron recordings during a probabilistic learning task shows that the human brain uses two separate signals to promote exploration under uncertainty. Neurons in the amygdala and temporal cortex change their firing before decisions to explore, and increased variability—or “noise”—in amygdala activity particularly drives exploration when people face potential losses.
Every day, humans and other animals must decide whether to exploit what is known or explore new options that might yield better outcomes or help avoid threats. This exploration–exploitation dilemma has been well studied in reward-seeking contexts, but far less is known about how exploration operates when avoiding losses. The new findings clarify how single neurons contribute to these decisions and how the brain flexibly shifts strategies depending on whether choices involve gains or threats.
Single-Neuron Signals During Decision-Making
Researchers recorded single-neuron activity from people performing a task that mixed gain and loss trials in a probabilistic learning environment. Participants repeatedly chose between familiar options and alternatives, creating situations that required either exploitation or exploration—similar to choices we face in everyday life.
Analyses showed that neurons in both the amygdala and the temporal cortex consistently altered their firing patterns just before participants chose to explore. This anticipatory modulation indicates that exploration is not purely random; instead, it is prepared by specific neural signals that bias the brain toward trying new options.
Loss Contexts Increase Exploration
Behaviorally, participants were more likely to explore when trying to avoid losses than when pursuing gains. At the neural level, this increased exploration was associated with a rise in the variability—or noise—of amygdala firing. In neuroscience, “noise” refers to spontaneous fluctuations in neuronal activity; here, greater noise reduced predictability and appeared to push choices toward riskier, exploratory behavior.
This pattern suggests an intrinsic bias in decision systems: when the stakes include potential losses, the brain becomes more willing to take chances to escape negative outcomes.
Two Mechanisms That Drive Exploration
The study identifies two complementary mechanisms that guide exploration:
- Valence-independent rate signal: A general preparatory change in firing rate that occurs before exploration regardless of whether outcomes are gains or losses. This signal appears across brain regions and primes neurons to enter an exploratory mode.
- Valence-dependent noise signal: A global increase in variability in amygdala activity that is specific to aversive or loss contexts and that promotes risk-taking and exploration.
Together, a baseline readiness to explore (rate) and a context-sensitive push toward variability (noise) allow the brain to adjust exploration dynamically based on whether decisions involve potential rewards or threats.
Relevance to Mood and Psychiatric Disorders
The results offer a mechanistic link between heightened amygdala activity observed in mood and anxiety disorders and altered exploration behavior seen in these conditions. If amygdala noise is amplified in people with anxiety or depression, it could drive maladaptive decision patterns—either excessive, destabilizing exploration or pathological avoidance—depending on how these signals interact with other neural systems.
By pinpointing the single-neuron correlates of exploration in both positive and negative contexts, the study provides a cellular-level framework to understand and eventually target decision-making abnormalities in psychiatric illness.
Looking Ahead
Future research should test whether modulation of these neural dynamics—via targeted neuromodulation, behavioral training, or pharmacology—can rebalance exploration and exploitation in individuals whose decision processes are disrupted. The findings also underscore the importance of studying both reward-seeking and loss-avoidance when investigating cognitive function and designing interventions for psychiatric disorders.
More broadly, by revealing that the amygdala’s variability helps shape choices under threat, the study reframes exploration as not only curiosity-driven but also a survival-oriented strategy tuned by emotional and valence-specific signals.
About this neuroscience research news
Author: Neuroscience News Editorial Team
Contact: Neuroscience News
Source: Neuroscience News Editorial Team
Image: The image is credited to Neuroscience News
Original Research: Closed access.
“Rate and noise in human amygdala drive increased exploration in aversive learning” by Tamar Reitich-Stolero et al., Nature
Abstract
Rate and noise in human amygdala drive increased exploration in aversive learning
To succeed in uncertain environments, animals must balance using current resources with searching for better alternatives. While exploration has been widely studied in reward-driven tasks and linked to frontal and subcortical structures, its single-neuron mechanisms in loss contexts remained unclear.
This study examined single-neuron dynamics while participants performed a probabilistic task with intermixed gain and loss trials. Neurons in the amygdala and temporal cortex modulated firing before exploratory decisions in both contexts. Humans explored more when avoiding losses, and increased noise in amygdala neurons contributed to that behavior.
Overall, human exploration appears to be driven by a valence-independent rate signal and a valence-dependent global noise signal, suggesting a neural basis for the altered exploration rates observed in mood disorders and their related maladaptive behaviors.