(This article is for autistic adults, their families, and professionals seeking to understand prediction error in autism. We will explore what prediction error is, how it affects autistic individuals, and practical strategies for coping. Understanding prediction error can help reframe daily challenges and support self-compassion, offering a new perspective on why ordinary surprises can feel overwhelming and how to navigate them more effectively.)

The plan was simple. You were going to stop at the store on the way home, pick up the two things on your list, and be in the door by six. Then the store had rearranged the aisles. The thing you needed wasn’t where it had been for the last four years. You stood in the new layout for what felt like a long time, trying to figure out where to go next, and by the time you got home you were not just late — you were undone. Hours later you were still off, and you couldn’t quite explain why a grocery store had wrecked your evening.

If this is familiar, it isn’t a sign that you’re fragile or overreacting. You’re likely experiencing what neuroscience calls prediction error in autism — and for autistic adults, the way the brain handles prediction errors is part of the daily story of why ordinary surprises sometimes cost so much.

This is one of the central frameworks in Sarah Bergenfield and Martha Sweezy’s recent book, which reframes autism as a condition of sensory surplus rather than deficit. Sensory surplus refers to the autistic brain receiving and processing more sensory information than neurotypical brains, which can lead to a perception of the world as unpredictable and overwhelming. Bergenfield, an autistic psychologist who was diagnosed at age 50, and Sweezy, a Harvard-affiliated psychotherapist, draw on neuroscience, polyvagal theory, and internal family systems to describe what it actually feels like to live inside a brain that’s working overtime to anticipate the world.

In this blog post, you’ll learn what prediction error is in plain language, how it tends to land differently in autistic nervous systems, why a misaligned prediction can produce a response that looks much bigger than the trigger, and what kinds of small adjustments — practical and self-compassionate — can help. The goal isn’t to fix anything. It’s to give you a clearer name for something you’ve probably been navigating, without language, for most of your life.

3 Key Takeaways

The brain predicts what’s coming next at every moment, and a prediction error is the signal it generates when reality doesn’t match the prediction.
For many autistic adults, prediction errors are louder, more frequent, and metabolically more expensive — which is why small surprises can feel disproportionately large.
Understanding prediction error doesn’t make the surprises stop, but it shifts the frame from “what’s wrong with me” to “what is my brain doing, and what does it need.”

What Prediction Error in Autism Actually Is

Most people picture the brain as a thing that reacts to the world — something happens, the brain registers it, you respond. Modern neuroscience tells a different story. The brain is less a reactor and more a forecaster. At every moment, it’s generating predictions about what’s about to happen next, based on every similar moment it has ever experienced. Then it compares those predictions to the sensory data coming in.

Prediction error refers to the brain’s signal when reality does not match its prediction. When the predicted outcome and the reality line up, the brain stays quiet. You don’t notice your brain doing this work, because nothing needs your attention. When they don’t line up — when the world doesn’t behave the way your brain predicted—a violation of expectation—the brain generates a prediction error. That signal is the brain saying: something here doesn’t match my map. Pay attention.

The predictive impairment hypothesis suggests that difficulties in social interactions and communication in autism may stem from challenges in estimating conditional probabilities related to social cues and behaviors. This means that autistic individuals may find it difficult to predict what will happen next in social situations, leading to increased anxiety and a reliance on routines or rituals as coping mechanisms.

Bergenfield and Sweezy, in their Psychology Today writing on this framework and hypothesis, describe the brain’s options at that point: it can treat the mismatch as noise and ignore it, update its model, or act on the world to make reality match the expectation. Most of the time, in most brains, this happens fast and quietly. The brain updates, the body settles, life continues. Psychology Today

Why the Brain Bothers Predicting at All

Prediction is metabolically efficient. If your brain had to process every incoming signal as if it were brand new, you’d be exhausted before breakfast. By predicting most of what’s about to happen, the brain saves enormous energy for the moments that actually need attention. Various factors—such as prior experience, sensory sensitivities, and environmental stability—influence how efficiently the brain can predict and process sensory input. The work of researchers like Lisa Feldman Barrett has been central to this view — the brain isn’t reacting to the world, it’s continuously building and updating a model of it, and only flagging the parts that don’t fit.

A small amount of prediction error is actually useful. It’s how we learn. A world with zero surprises offers nothing new to update toward. A world with constant surprises offers too much to process. The nervous system tends to function best somewhere in between, in what Bergenfield and Sweezy describe as the range where there’s enough uncertainty to stimulate learning, but not so much that the system shifts into protection and overwhelm—the extent of uncertainty that stimulates learning without overwhelming the system. Psychology Today

How Prediction Error in Autism Tends to Land Differently

Here’s where the picture gets specific to autistic individuals and how their brains differ in processing prediction errors. There’s a growing body of research — and the framework Bergenfield and Sweezy build their book around — suggesting that autistic brains process prediction errors differently. The signals tend to land louder, more frequently, and with less filtering than the same signals in a non-autistic brain. Research provides a consistent account of these differences in prediction error processing, helping to explain the unique cognitive experiences of autistic individuals.

Sensory Surplus and Prediction Error

This is connected to what they call sensory surplus. Sensory surplus refers to the autistic brain receiving and processing more sensory information than neurotypical brains, which can lead to a perception of the world as unpredictable and overwhelming. Where a non-autistic brain might filter out a lot of incoming information before it ever reaches conscious attention, an autistic brain often receives much more of it, with a heightened focus on sensory processing differences. More sound. More light. More texture. More social signal. More micro-detail of every kind. All of that increases the raw amount of sensory data that needs to be matched against prediction at every moment within a broader neurodivergent way of experiencing the world. Which means more potential places for a mismatch. Which means more prediction errors, and highlights how prediction error processing can differ between autistic and non-autistic individuals.

Daily Life Examples

Many children on the autism spectrum, as well as adults, often display a strong insistence on sameness and resistance to change—behaviors that serve as coping mechanisms for unpredictability and anxiety in daily life. Here are some examples of how prediction error can manifest:

A small change in routine — a different route, a moved object, a meeting time pushed by fifteen minutes — producing a response that feels far out of proportion to the change.
An unexpected sound, smell, or texture cutting through whatever you were doing and taking minutes or hours to recover from.
Walking into a familiar room and feeling something is “off” before you can identify what’s different.
A conversation taking a turn you didn’t anticipate, and finding yourself unable to follow the new direction even though you can usually keep up.
Coming home from a day that “shouldn’t” have been hard and discovering you have nothing left, because the accumulated prediction errors added up.

These behaviors—such as insistence on sameness and resistance to change—are not just quirks; they are deeply linked to a strong need for predictability and help mitigate anxiety that arises from unpredictable sensory environments.

None of these are signs of being fragile. They are signs of a system that is processing more, predicting more, and getting hit by more mismatches than the system around it.

Coping Mechanisms

Relying on routines and rituals to reduce unpredictability.
Seeking out familiar environments, foods, and objects.
Avoiding situations with high sensory or social unpredictability.
Taking deliberate recovery time after unexpected changes or high-prediction-error events.

Why This Is Metabolically Expensive

Every prediction error costs energy to process. The brain has to decide what to do with the mismatch — ignore it, update for it, or act on it. In an autistic system with sensory surplus, those decisions are happening far more often than in a non-autistic system. The cumulative cost isn’t theoretical. It shows up as exhaustion, irritability, sensory shutdown, or the kind of evening collapse that follows a day that looked, from the outside, completely manageable.

This is part of why autistic adults often need more recovery time than the day’s events seem to warrant, and why experiences like autistic burnout can emerge when that recovery never quite catches up. The visible event isn’t the whole story. The invisible story is how many prediction errors the brain processed to get through it.

Neural Correlates of Brain Function

Understanding the neural patterns that shape how autistic brains experience the world offers profound clarity about why prediction and sensory experiences feel so uniquely different for those of us on the spectrum. When we talk about “neural correlates,” we’re simply naming the specific brain regions and activity patterns that create our inner landscape—how we think, feel, and respond to everything around us. In autistic nervous systems, research consistently reveals differences in both brain structure and function, particularly in areas that navigate social connection, communication, and the intricate dance of sensory processing.

Predictive Coding Regions

Here’s what matters most: the brain regions responsible for predictive coding—our internal system for anticipating what comes next—operate with their own beautiful logic in autism. The anterior cingulate cortex and insula, two regions that typically detect mismatches between expectation and reality, show altered activity patterns in autistic individuals. This isn’t broken—it’s different. These areas may process information with less automatic filtering, making it challenging to dismiss irrelevant sensory input or adapt swiftly to unexpected shifts. This explains why even small surprises or routine changes can feel overwhelming to the nervous system. The neural pathways that might usually soften or resolve prediction errors are simply following a different rulebook.

Implications for Support

These differences aren’t deficits to fix—they’re features of a brain brilliantly tuned to process more detail, more intensely, often with exquisite sensitivity. This neural wiring creates both extraordinary strengths—like detecting patterns others completely miss—and genuine challenges, such as navigating communication or adapting to new environments. When we truly understand these neural patterns of autism, we gain clarity about why the spectrum is so beautifully vast, and why support must honor each person’s unique neurological blueprint, whether or not someone fits labels like “high-functioning” autism. Current research continues deepening this understanding, offering genuine hope for approaches that are both effective and deeply respectful of autistic lived experience.

Why Predictability Matters So Much

Once you understand prediction error, a lot of things start to make sense. Including the things that often get pathologized — the routines, the rituals, the strong preference for familiar foods and clothes and routes and people. These aren’t quirks. They aren’t rigidity. They are, in nature, a nervous system intelligently reducing the number of prediction errors it has to process in a day.

When routines are disrupted or the environment becomes unpredictable, the possibility of prediction errors increases, making daily life feel more overwhelming or unsafe.

Why Routine Works

A familiar route home means your brain doesn’t have to predict and re-predict at every turn. A familiar meal means your senses know what’s coming. A familiar piece of clothing means your skin isn’t constantly flagging new texture against your nervous system. None of this is avoidance. It’s bandwidth management. The brain is reserving its prediction-error budget for the parts of the day that genuinely require it — work, relationships, novelty by choice — instead of spending it on a coat that itches.

This is also why disruptions to routine can feel so disproportionate. From the outside, “the cafe was closed” is a minor inconvenience. From the inside, it’s a cascade of prediction errors — the planned route is gone, the planned drink is gone, the planned ten-minute pause is gone, and now the brain has to generate new predictions for the next half hour while the body is still settling from the original mismatch.

What This Can Look Like

Eating the same breakfast for years, not out of inflexibility, but because that’s one fewer thing the brain has to predict each morning.
Preferring to drive a familiar route even when a faster one is available, because familiarity is a real resource.
Needing more warning than other people about schedule changes — not because change is impossible, but because the brain needs time to update its model.
Becoming attached to specific objects, clothes, or places that have proven low-prediction-error over time.
Finding new environments genuinely exhausting in a way that’s not really about social anxiety — it’s about the volume of unpredicted sensory data.

None of these patterns are problems to solve. They are sensible adaptations to a system that processes a lot.

Hyperplasticity and Brain Development

The Role of Hyperplasticity in Autism

Hyperplasticity speaks to the autistic brain’s extraordinary gift for reshaping itself in response to the world around us. When we look through this lens at the autistic experience, we begin to understand why certain ways of being—like the need for repetition and the intense dance with sensory overwhelm—feel so essential to who we are. The autistic nervous system moves through periods of heightened adaptability, especially in those tender early years, creating a rich tapestry of neural pathways that wire us differently from the ground up.

This hyperplastic nature means our brains drink in sensory information with an intensity that can feel like standing under a waterfall when all we needed was a gentle rain. For so many of us living this experience, the world can feel like it’s coming at us in waves that crash without warning, making routine and familiar patterns not just comforting but necessary—they become our anchor in the storm. When researchers study animal models that mirror our neurological patterns, they see the same intricate rewiring, the same beautiful complexity that validates what we’ve always known about our own lived truth.

Here’s what matters most: this hyperplasticity isn’t something that needs fixing. The same nervous system sensitivity that can leave us feeling raw and overwhelmed also holds space for profound learning and adaptation that flows in ways others might never experience. The relationship between our plastic, responsive brains and autism is beautifully complex, and each day brings new understanding about how these neurological gifts shape the full spectrum of who we are. When we honor hyperplasticity for what it truly is, we can hold both the tender challenges and the remarkable possibilities that come with an autistic way of moving through the world—and build support that doesn’t try to change us, but helps us thrive exactly as we are.

What Helps: Working With the Predictive System, Not Against It

The instinct, especially for late-identified autistic adults who spent decades being told to be more flexible, is to try to override these patterns. Push through. Be more adaptable. Stop being so rigid. This rarely works, because the patterns aren’t choices. They’re how the system manages its own load. There are important limitations to how much we can override the brain’s prediction-error management strategies; these are deeply rooted processes, not simply habits that can be changed at will.

The work that does tend to help looks less like overriding and more like cooperating. Building a life that respects the brain’s actual prediction-error budget. Reducing avoidable mismatches. Recovering deliberately from unavoidable ones. Bergenfield and Sweezy’s work specifically points to internal family systems and similar approaches as ways to do this work compassionately — by getting curious about the system’s responses rather than treating them as failures.

Why Cooperation Works

When the nervous system learns that prediction errors will be met with care rather than pressure, it tends to stay in its working range more often. The protective responses — shutdown, meltdown, sensory withdrawal — get triggered less because the system isn’t constantly braced for a fight with itself. This doesn’t make the underlying wiring different. It makes the wiring’s daily experience less hostile, which has real downstream effects on what the system can actually do.

What This Can Look Like in Practice

Building in predictability where it’s free to build. Same morning routine, same workspace, same lunch — not because variety is bad, but because saving prediction-error budget here means you have it for the parts of the day that need it.
Warning the brain before changes. Even a few minutes of mental preparation before a known disruption — a different route, an unfamiliar meeting, a doctor’s appointment — gives the predictive system time to update its model in advance.
Recovering on purpose after high-prediction-error events. A new environment, a social event, a change in plans, a travel day — all of these spend more budget than they look like they should. Planning for recovery isn’t weakness. It’s accurate accounting.
Reducing avoidable sensory load. Clothes that don’t fight your skin, environments that don’t fight your ears, lights that don’t fight your eyes. These small reductions free up budget for the prediction errors you can’t avoid.
Naming what happened, without shame. “That was a lot of prediction error for my system” lands differently than “I overreacted.” The first is information. The second is shame.

What This Looks Like in Relationships

If you live with someone, the prediction-error framework changes what counts as a meaningful change of plans. A last-minute switch that costs your partner a moment of mental flexibility might cost you a much larger amount of nervous system budget, especially in neurodiverse relationships where each person’s sensory and predictive systems work differently. This isn’t because you’re more sensitive. It’s because the same change runs through a different system. Naming this out loud — “That kind of change costs me more than it looks like it should” — can replace a lot of conflicts about why you “can’t just go with the flow.”

When Prediction Error and Autism Meet Late Diagnosis

For autistic adults who weren’t identified until adulthood, learning about prediction error often produces a particular kind of recognition. You realize that the responses you’ve spent your life apologizing for — the freezes at small changes, the disproportionate fatigue, the need for routine, the after-effects that lingered for days — were never personality flaws. They were a nervous system telling the truth about how much it was processing.

This is one of the gifts of the late-identified journey. The framework doesn’t change what happened in your past. But it changes the meaning. The decades of being called too rigid, too dramatic, too sensitive, too much — those weren’t accurate readings. Those were people without the language for what your system was actually doing, including the ways neurodivergent masking can hide distress in plain sight, especially for autistic women and femmes.

It also tends to soften the shame. When you can see that a difficult evening was the predictable downstream cost of a high-prediction-error day, you stop having to explain it to yourself as personal failure. You start being able to plan around it, recover from it, and let it be what it is. Bergenfield writes openly about her own late diagnosis at 50, and the way that recognition reshaped her relationship to her own responses. For many, weaving these insights into a new self-story through approaches like narrative therapy for autistic adults can be a powerful part of that reframing. It’s a path many late-identified adults walk.

Integration: What to Take From This

If you take one thing from this, let it be this: when small things feel big, your brain isn’t malfunctioning. It’s doing exactly what it has always done — predicting the world, comparing reality to the prediction, and generating signals when the two don’t match. The identification of prediction error is a key step in understanding your own experience, as it helps clarify what is happening internally. These processes are not unique to autism; they are fundamental to human brain function, and prediction error is a universal part of how all human brains interpret and respond to the world. The difference for many autistic adults isn’t that this process is broken. It’s that more of the world produces mismatches, the mismatches are felt more intensely, and the cost of processing them is real.

That has practical implications. You probably need more predictability than the average person, and that’s not a failing — it’s an accurate response to how much you process. You probably need more recovery time, and that’s not laziness — it’s the visible cost of invisible work. You probably benefit from being able to name prediction error when it’s happening, because naming changes how the system carries it.

The work isn’t to become a brain that doesn’t generate so many prediction errors. That brain isn’t available to you, and pretending it is has been part of what wore you out. The work is to build a life that has room for the brain you actually have — the one that notices more, predicts more, costs more to run, and also, in many cases, sees and feels and connects more than the systems around it.

You don’t need a quieter brain. You need a life sized for the one you have.

A Soft Invitation

If something in this resonated and you’d like to talk it through, I offer a no-pressure conversation to see whether this kind of work might fit for you, whether you’re seeking therapy as an autistic or ADHD adult yourself or wanting support in communicating more effectively with autistic adults in your life.

Book a FREE 20-minute “Clarity & Connection” Zoom session

Frequently Asked Questions

Is prediction error a problem to fix?

No. Prediction error is a normal feature of every brain — it’s how brains learn. The difference for many autistic adults isn’t that prediction errors are happening; it’s that they happen more often, land louder, and cost more to process. Sometimes, the absence of prediction-error signals in neural data can indicate either a lack of prediction or highly variable neural responses, which complicates detection in research. The work isn’t to eliminate them but to build a life that respects how much processing the system is actually doing.

Does this mean autism is a brain disorder?

Bergenfield and Sweezy explicitly reframe it the other way. Their book describes autism as a condition of sensory surplus rather than deficit — a brain receiving and processing more rather than less. Prediction error is part of what that surplus costs. The temporal dynamics of prediction error processing in autism—how these signals unfold over time—are an important area of research, highlighting differences in timing and propagation of neural responses. The framing matters, because “disorder” tends to push toward fixing, while “different processing” tends to push toward accommodation and respect.

Why do small changes affect me so much when I can handle big stresses?

Big stresses are often expected. Your predictive system has time to build a model for them. Small changes — a moved object, a different route, an unexpected sound — bypass that preparation and produce raw prediction errors with no warning. The size of the trigger isn’t what determines the cost. The amount of mismatch, and whether the system had time to update for it, is what determines the cost.

How does this relate to sensory overload and meltdowns?

Meltdowns and shutdowns are often what happens when the cumulative prediction-error budget runs out. Each individual mismatch may have been manageable, but together they exceed what the system can process, and the protective responses come online. Research has identified precision weighted prediction errors as key neural signals involved in these processes, reflecting how the brain dynamically weights the importance of prediction errors in sensory and social situations. Understanding this can help with planning — building in recovery time before the budget runs out, rather than trying to push through a system that’s already over capacity.

Where can I learn more about this framework?

Wired to Feel: Autism as a Condition of Sensory Surplus by Sarah Bergenfield and Martha Sweezy is the primary book-length treatment of these ideas. Martha Sweezy has also written about the predictive brain and internal family systems on Psychology Today, including pieces co-authored with Bergenfield. Lisa Feldman Barrett’s work on predictive processing offers a broader scientific foundation if you want to go deeper into the neuroscience underneath. Animal models, such as VPA-treated marmosets, have been used to study prediction error mechanisms relevant to human autism, providing valuable insights into neural processes. Researchers have also identified neural markers for autism—measurable brain features associated with predictive coding and sensory processing—which are used in both research and diagnosis. None of these are replacements for working with someone who knows your specific system, but they’re a good starting place.

Prediction Error in Autism and the Autistic Brain: Why Small Surprises Cost So Much