Neurocomputational evidence of sustained Self-Other mergence after psychedelics

This study demonstrates that psychedelics (psilocybin and 2C-B) induce a sustained state of Self-Other mergence by reducing inhibitory neural connectivity from the right temporoparietal junction to the dorsomedial prefrontal cortex, a neurocomputational change associated with improved psychosocial wellbeing.

The Gist

The Big Idea: Blurring the Lines Between "Me" and "You"

Imagine your mind is a house with very sturdy walls. Inside your house is You, and outside is Everyone Else. Usually, these walls are essential. They help you know what you think, feel, and believe, separate from what your neighbor thinks, feels, and believes. This separation is called Self-Other distinction.

However, sometimes these walls can get a little too rigid, making it hard to connect with others or understand their perspective. Conversely, if the walls are too thin, you might get confused about whose thoughts are whose.

This study investigated what happens to these "walls" when people take psychedelic drugs (specifically psilocybin and 2C-B). The researchers wanted to know: Do these drugs tear the walls down? And if so, does that help people feel more connected and happy?

The Experiment: A Game of "Guess the Umbrella"

To test this, the researchers didn't just ask people how they felt. They gave them a tricky brain game called the Probabilistic False Belief Task.

The Analogy:
Imagine you are watching a tourist shop on a tropical island.

• You see a sequence of items being sold: sometimes pink umbrellas, sometimes yellow sun-shades.
• There is also a Shop Manager (a fictional character) who is watching the shop too, but sometimes he is in a different room, and sometimes he is looking at a broken security camera that shows old footage.

The Challenge:
You have to keep track of two things at the same time:

1. What is actually happening in the shop right now? (Your belief).
2. What does the Manager think is happening, even if he is looking at the wrong info? (His belief).

Usually, your brain keeps these two tracks separate. You know, "I know it's an umbrella, but he thinks it's a sun-shade."

The Result:
When the participants took the psychedelics (and were tested the next day), their brains started to mix the tracks. They began to update their own beliefs based on what the Manager saw, and vice versa. The "walls" between their mind and the Manager's mind got leaky.

In scientific terms, the drugs caused Self-Other Mergence. The participants didn't get bad at the game; they just stopped treating the Manager's thoughts as completely separate from their own. They started to feel like, "If he sees it, I see it too," even when they knew he was looking at the wrong screen.

The Brain Scan: The "Security Guard" Loosens Up

To understand why this happened, the researchers looked at the participants' brains using a super-powerful MRI scanner (7 Tesla) while they were under the influence of the drugs.

The Analogy:
Think of your brain as a busy office building.

• The Right Temporoparietal Junction (rTPJ) is like a Security Guard. Its job is to stand at the door and say, "Stop! That thought belongs to the Manager, not to You!" It puts up an inhibitory (stopping) signal to keep your thoughts separate from others'.
• The Dorsomedial Prefrontal Cortex (dmPFC) is the Manager's Office where all the thinking happens.

The Discovery:
The brain scans showed that the psychedelics weakened the Security Guard's signal. The guard stopped shouting "Stop!" as loudly. Because the guard was less strict, the flow of information between "You" and "The Manager" became freer. The "leak" in the belief system was caused by this specific connection in the brain losing its grip.

The "Afterglow": Why Does This Feel Good?

You might think, "If I can't tell my thoughts from others, that sounds confusing or scary!" But the study found something surprising.

The participants who showed the most "leakiness" (the most merging of Self and Other) the day after taking the drug also reported feeling the most positive. They felt more connected to humanity, more empathetic, and generally happier.

The Analogy:
Think of the "Self-Other wall" as a fence in a garden.

• Normal state: The fence is high and locked. You are safe, but you can't see the neighbors' flowers, and they can't see yours.
• Psychedelic state: The fence is lowered. You can see the neighbors' flowers, and they can see yours. You feel a sense of shared beauty and connection.

The study suggests that for a short time, lowering this fence allows for a "re-learning" of how we connect with others. It's like a reset button for our social brains, allowing us to practice being more open and less isolated.

The Takeaway

1. It's not confusion; it's connection: The drugs didn't make people "dumb" or unable to tell the difference between reality and fantasy. Instead, they specifically changed how the brain processes other people's minds, making them feel less separate.
2. The brain has a specific switch: This effect is linked to a specific pathway in the brain (the Security Guard to the Manager's Office) that usually keeps us separate. Psychedelics temporarily turn down the volume on that switch.
3. Therapeutic potential: Many mental health issues (like depression or loneliness) involve feeling stuck in our own heads, separated from the world. By temporarily softening these boundaries, psychedelics might help people "re-learn" how to connect, leading to lasting feelings of well-being and empathy.

In short, the study shows that psychedelics don't just break your mind; they temporarily dissolve the walls between "Me" and "You," and for many, that shared space feels like a profound relief.

Technical Summary

1. Problem Statement

Mental illnesses such as schizophrenia, autism, and borderline personality disorder are often characterized by maladaptive "Self-Other" distinctions (the ability to differentiate one's own mental states from those of others). While serotonergic psychedelics (e.g., psilocybin, 2C-B) are known to induce "ego dissolution" and enhance social connectedness, the neurocomputational mechanisms underlying these changes remain unclear.

Existing research relies heavily on subjective reports. There is a lack of objective, quantifiable behavioral biomarkers that can measure how psychedelics alter the fundamental computational processes of belief formation and learning, specifically regarding the distinction between "Self" and "Other" agents. This study aims to bridge that gap by applying computational psychiatry frameworks to determine if psychedelics induce a sustained state of "Self-Other mergence" in belief updating and to identify the neural connectivity changes driving this effect.

2. Methodology

Study Design:

• Type: Double-blind, placebo-controlled, within-subject crossover design.
• Participants: $N=22$ healthy volunteers (11 female, aged 19–35) with prior psychedelic experience but no use in the last 3 months.
• Conditions: Participants received three doses separated by 2-week washout periods:
  1. 15 mg Psilocybin
  2. 20 mg 2C-B (2,5-dimethoxy-4-bromophenethylamine)
  3. Placebo
• Timeline:
  • Day 0 (Acute): 7T resting-state fMRI (rsfMRI) acquired ~110 minutes post-administration. Cognitive tasks (Tower of London, DSST, Spatial Memory) administered during peak effects.
  • Day 1 (Subacute): Probabilistic False Belief Task (pFBT), Stroop task, and Visuospatial Perspective Taking (VPT) task. Subjective questionnaires (PANAS, IOS, SCS-R).
  • Day 5 (Subacute): Persisting Effects Questionnaire (PEQ) and follow-up subjective measures.

Behavioral Assay: Probabilistic False Belief Task (pFBT)

• Task: Participants tracked a drifting probability ($P$) of an outcome (e.g., pink umbrella vs. yellow sun-shade). They also had to track a fictional "Other" agent's belief about $P$.
• Conditions:
  • Shared: Both Self and Other see the outcome.
  • Privileged: Only Self sees the outcome (Other is unaware).
  • Decoy: Other sees misleading/outdated information.
• Goal: Participants reported their own belief ($P_{Self}$) and the Other's belief ($P_{Other}$) on probe trials.
• Computational Modeling:
  • A family of 36 Reinforcement Learning (RL) models was fitted to the data.
  • Key Parameter ($\Lambda$): A "Self-Other leak" parameter quantifying the extent to which prediction errors (surprise signals) from one agent erroneously update the belief of the other.
  • $\Lambda = \alpha \times |\lambda|$, where $\alpha$ is the learning rate and $\lambda$ is the leak coefficient.
  • Hypothesis: Higher $\Lambda$ indicates Self-Other mergence (agent-independent prediction errors), while $\Lambda \approx 0$ indicates strict Self-Other distinction.

Neuroimaging Analysis:

• Modality: 7 Tesla resting-state fMRI.
• Regions of Interest (ROIs): Five core Theory of Mind (ToM) nodes: dmPFC, vmPFC, Precuneus, lTPJ, rTPJ.
• Analysis: Spectral Dynamic Causal Modeling (sDCM) to estimate Effective Connectivity (EC) (directed influence between regions).
• Statistical Framework: Parametric Empirical Bayes (PEB) to model group-level connectivity and its association with the behavioral parameter $\Lambda$.

Control Measures:

• To rule out general cognitive impairment as a confound, the study included the Stroop task (response conflict) and VPT (egocentric conflict) on Day 1.
• An "impairment burden" composite score was created to test if changes in $\Lambda$ correlated with general cognitive decline.

3. Key Results

A. Behavioral Findings: Sustained Self-Other Mergence

• Task Performance: Psychedelics did not impair overall task accuracy. Participants maintained the ability to track probabilities, indicating the effect was not due to general cognitive failure.
• Model Comparison: A 3-parameter model (Learning Rate $\alpha$, Decision Temperature $\tau$, Self-Other Leak $\Lambda$) best explained the data.
• Drug Effect: Both Psilocybin and 2C-B significantly increased Self-Other mergence ($\Lambda$) compared to placebo in the subacute phase (Day 1).
  • Psilocybin: $p = 0.029$
  • 2C-B: $p = 0.21$ (trend, but mean increase comparable to psilocybin).
  • Combined Drug vs. Placebo: $p = 0.02$.
• Specificity: The increase in $\Lambda$ was specific to belief updating. There were no significant drug effects on learning rate ($\alpha$) or decision temperature ($\tau$), suggesting the effect is not merely increased sensitivity to surprise or random guessing.
• Cognitive Confounds: No correlation was found between the increase in Self-Other mergence and the "impairment burden" composite. Participants actually showed improved interference control on the Stroop task post-drug, ruling out general executive dysfunction as the cause.

B. Subjective Correlates

• Psychosocial Wellbeing: Increased Self-Other mergence ($\Lambda$) was positively correlated with a multivariate profile of psychosocial wellbeing (including ego dissolution, altruistic prosocial effects, and positive affect) measured acutely and subacutely.
• Canonical Variate: Individuals with higher $\Lambda$ reported greater "afterglow" effects, suggesting the computational shift underpins the therapeutic "connectedness" experience.

C. Neuroimaging Findings: rTPJ $\to$ dmPFC Disinhibition

• Baseline: In the placebo state, there was a strong inhibitory connection from the right Temporoparietal Junction (rTPJ) to the dorsomedial Prefrontal Cortex (dmPFC).
• Drug Effect: Psychedelics induced a significant attenuation of this inhibitory tone (relative disinhibition).
• Behavioral Link:
  1. Participants with weaker baseline rTPJ $\to$ dmPFC inhibition showed larger drug-induced increases in Self-Other mergence.
  2. The magnitude of the drug-induced disinhibition predicted the magnitude of behavioral Self-Other mergence 24 hours later.
• Specificity: This effect was specific to the ToM network; control analyses using the Salience/Ventral Attention network did not show similar associations.

4. Key Contributions

1. Objective Biomarker: The study provides the first objective, computational metric ($\Lambda$) for quantifying "Self-Other mergence" in belief formation, moving beyond subjective reports of ego dissolution.
2. Mechanism of Action: It identifies a specific neurocomputational mechanism: psychedelics reduce the agent-specificity of prediction errors, allowing "leakage" of surprise signals between Self and Other representations.
3. Neural Circuitry: It pinpoints the rTPJ $\to$ dmPFC pathway as the critical neural substrate. The rTPJ normally inhibits the dmPFC to maintain distinct agent perspectives; psychedelics relax this inhibition, facilitating a merged state.
4. Therapeutic Relevance: The findings link the computational shift (mergence) directly to the "afterglow" period of psychosocial wellbeing, supporting the hypothesis that psychedelics reopen a critical period for social learning and cognitive restructuring.
5. Differentiation from Impairment: The study rigorously demonstrates that this mergence is a specific alteration in social cognition, not a byproduct of general cognitive impairment or executive dysfunction.

5. Significance and Implications

• Precision Psychiatry: The ability to quantify Self-Other mergence offers a potential biomarker for predicting individual responses to psychedelic-assisted therapy. It could help identify patients who may benefit from the "re-opening" of social boundaries versus those (e.g., those with Borderline Personality Disorder who already exhibit maladaptive mergence) who might be at risk.
• Refining the REBUS Model: The findings extend the REBUS (Relaxed Beliefs Under pSychedelics) model. While REBUS suggests a general reduction in prior precision, this study shows a specific relaxation of Self-Other boundaries in the meta-representation of prediction errors.
• Clinical Translation: Understanding that the rTPJ $\to$ dmPFC pathway is modifiable suggests potential adjunctive interventions (e.g., TMS) to modulate maladaptive belief updating in psychiatric disorders without the use of psychedelics.
• Theoretical Shift: The paper challenges the view that psychedelics only affect "affective" empathy, showing they also alter "cognitive" empathy (belief attribution) by blurring the computational boundaries between self and other.

In summary, this paper establishes that psychedelics induce a sustained, quantifiable state of Self-Other mergence in healthy adults, driven by the disinhibition of a specific cortical pathway (rTPJ to dmPFC), which correlates with enhanced psychosocial wellbeing and offers a new computational framework for understanding the therapeutic potential of these compounds.