A Complete Answer To The Tulpa Question

[Mai_x_v3]

I'm glad to see you back! You have such detailed progress reports I'm so sorry about you being sick and your family member having cancer, it's awful to have to go through that, nobody deserves that. I hope you all are doing well now! 💕🩷 Also I love all the extra details on your log. Good luck with your Tulpa forcing!

[Biscuit]

On 10/10/2025 at 1:59 PM, Mai_x_v3 said:

I'm glad to see you back! You have such detailed progress reports I'm so sorry about you being sick and your family member having cancer, it's awful to have to go through that, nobody deserves that. I hope you all are doing well now! 💕🩷 Also I love all the extra details on your log. Good luck with your Tulpa forcing!

Thank you for your kind words!!! All is going perfectly swell now luckily! Thank you for caring enough to read, and it's always a pleasure to see a fellow V3 fan out in the wild

Edited October 18, 2025 by Biscuit

[Biscuit]

Week 1

Day 2

 

     Didn't expect I would be back so fast! I know I said I'd post only when something happened I could report on, but I also wanted to extend that to posting when I have something important to say. I think it would be pretty novel to actually lay out a detailed model of Tulpas as I am right now, and see how this model changes when it comes into contact with my lived experience creating a Tulpa.  This model isn't based on nothing, obviously, but a good long while of observation and good ol' thinkin.

 

 

An extremely (perhaps PAINFULLY so) detailed breakdown of my best guess as to how Tulpas work is in the spoiler below, read at your own peril! (If you are a well seasoned Tulpa host and are up to the task, I would ask you to read it and challenge it to see where it fails in your experience! Pretty please!)

 

 

Spoiler

For the fun of it all, for the love of the game, let us attempt to build a hypothetical model of what a Tulpa is and how it works.

 

DISCLAIMER: This is nothing but thoughts and considerations!

 

 

 

Definition

 

 

The overall definition of a Tulpa, to me, is as follows: An autonomous system of thought contained within a living intelligence manifesting as an individual that possess an independent and self-maintaining, self-functioning ego.

 

The big takeaways here are these:

 

Autonomous - The Tulpa does not need the conscious piloting of the intelligence they inhabit to make decisions when impulsed to act.

 

Self-maintaining - The Tulpa does not require their host to consciously remember and administer aspects of their being to keep them together - they persist with or without the intervention of the intelligence they inhabit.

 

Self-functioning - The Tulpa does not need the conscious mind of the intelligence they inhabit to impulse them to act, they have their own mechanism to act when they are rendered.

 

 

Definitions of my own shorthand (Will be used later)

 

Tokens - Symbolic imagery that can easily bring up a complex and large network of thought through relationships to other symbols and thoughts. If we look at a picture of a person, their hair is a token that brings up memories and thoughts of people with that hairstyle, combine all the tokens and a basic rendering of how we imagine that person to be is there. Smell your childhood shampoo and it dredges up all the connected memories and feelings. These tokens can be constructed and manipulated at will, used in various studying techniques like the mind palace, and is a pretty well observed and understood thing to my knowledge.

 

Associative Symbol Network (ASN): The large mesh of data, symbols, and relationships between them that create the Memorized Model of a Tulpa. Allows quick recall of a Tulpa, and quick retrieval of information.

 

Empathetic Driver - A mechanism we all have that uses the pool of all memory and experience with people to simulate personhood. Present in a task as simple as imagining what your friend would do if you sneezed on them as the mechanism that is used to simulate that instantly. The empathetic driver being "encoded" simply refers to it mimicking a specific individual (like your friend who you sneezed on (why'd you do that))

 

 

 

 

 

How a Tulpa Works | Theory of Functionality

 

     A Tulpa's make up is a strongly reinforced mesh of tokens and relationships between thoughts we pass through an instantaneous semi-unconscious simulation of personhood driven by our empathetic understanding developed through our life (The empathetic driver extrapolating imminent behavior through the ASN). The marriage of this empathetic driver with the constructed mesh results in a highly sophisticated simulation* of a real life intelligence. The more we strengthen, explore, and build that network of tokens, thoughts, and connections the more consistent and realized this personhood becomes. The empathetic driver handles most of the weight of simulation by skipping the vast majority of steps and relying on a simple chain of cause and effect. If I poke you in real life, your nervous system feels the pressure, sends a signal, your brain fires, a lot of things that I am not NEAR qualified to explain happens, and you respond to both the sensation and the context of the situation. When I poke a Tulpa, my brain thinks about the type of person my Tulpa is and how they would react to the action of being poked and makes that happen, a different simulation entirely. When this simulation is entirely automated, we can reckon the Tulpa is their own person.

 

[NOTE: Think of the word simulation here not as This thing is not real and more of The intelligence of the host literally has to simulate existence]

 

 

 

 

 

 

How a Tulpa Exists in a Mind | Theory of Rendering

 

Tulpas are made of thought and are not subject to the rules of material reality, and as such their consistently maintained presence is not required to constitute a full existence, and from what I can gather from the consensus of neuroscience that would be impossible.

 

     Let us say a Tulpa can exist in two states, an Active Model and Memorized Model. An active model draws upon what is stored in the memorized model and uses a model of autonomy to bring the Tulpa to life through a projection into the mind's eye. Changes that occur are stored in the memorized model. A Tulpa is an interplay of the memorized model being pinged, impulsed, and activated by the incidental mechanisms of the host's mind, and projected into the brain. Instead of thinking of the Tulpa as one bundle of information in the Mind's Eye, we have to instead think of them spread out across memory and mechanisms. In essence - a Tulpa's existence is a holographic projection from a Memorized Model, with a living autonomous mechanism ensuring that process. The chain of mechanisms that makes actions possible and seeable is called the Active Model.

 

 

 

     A Tulpa cannot do anything if not impulsed to act. In a developed Tulpa impulses are automatically generated in any number of ways (Tokens/Stimulus remind the brain to "ping" the Tulpa, the host remembers the Tulpa/brings them into attention, in advanced cases perhaps it is entirely unconscious but this requires more exploration), but when impulsed to act the relevant model of the Tulpa is brought from memory and into the imagined "reality". It stands to reason that the entire model does not need to be activated at once, and that perhaps simpler interactions beget a smaller activation of the larger model.

 

     In this way, Tulpa that are not either doing something or having the host's attention (they are "forgotten" in the same way that you don't think about your phone beside you on your desk when doing a task), they are stored in memory and don't inhabit any other place. When a Tulpa does something outside the view of the host, the memorized model simulates and changes, not an active model, and this is likely "catch-up" done when the host interacts with the Tulpa after a gap of non-interaction. (Important to note this is functionally identical to a model of the classic persistent Tulpa in all forms of personhood and autonomy, just a different methodology of function.) Though I think this fails to explain how people can have their Tulpas be silent when calling to them, and have to "find" them. This is a behavior that has nearly been nearly unilaterally experienced, so that needs considering.

 

[Chart]

Spoiler

 

 

 

How a Tulpa Thinks | Theory of Thought

 

     In my proposed model, a Tulpa thinks by unconsciously making the host access the Memorized Model, utilize the encoded Empathetic Driver to receive a predicted Imminent Behavior, and make that happen in the mind's eye. To me there is no silver bullet that can make a Tulpa work in the same way as our Self does (in regards to the mechanical function). Our primary ego is the default response to stimulus and context, while any other ego is a predicted response with understanding of someone's perceived mind. We can weaponize this in the case of Tulpa's because our perceived mind is identical to the actual mind, so the simulation is 100% accurate.

 

     At the current time, I have no concrete mechanism as to how Tulpa grow in unexpected ways (another nearly unilateral experience) and become their own person far beyond what their host intended. The only thing I can think of is a bit of random chaos from intrusive thoughts can make something unexpected happen, and we interpret it as an aspect of character we didn't understand, and the Tulpa can use this to express novel ideas and thoughts. I am not very satisfied with this, though. The process seems too intentional.

 

How a Tulpa is a Self | Theory of Autonomy

 

How a Tulpa gains autonomy is extremely tricky. I wrote an article (for myself (yes, I'm beyond hope)) a bit ago about the problems with understanding Tulpa Autonomy, I'll put it in a spoiler below, but it will help us be on the same page.

 

Spoiler

Defining Tulpa: Resolutions of Autonomy

 

 

There is a fundamental issue at the heart of creating a Tulpa that has haunted the community for as long as I have known it: There is no proof of autonomy that is guaranteed.

 

Now I'm not harping on the miserably tried and true "Prove to me Tulpa's are real." We are all well aware that everything we know can only be based on testimony and deduction. What I'm trying to get at is one cannot know the degree of one's Tulpa's true autonomy with empirical evidence, which leads to fears that you are deluding yourself into having a mundane imaginary friend you think is a Tulpa. Now as a community we have largely moved on from this fear because it tends to cause nothing but problems, and people that do "jump the gun" on believing their Tulpa's are sapient/autonomous early end up getting them there through interaction anyway.

 

The problem still hangs rather annoyingly, though.

 

I have tried to solve this with my own methodology of Tulpa creation by wielding a constant doubt that hovers over every step of the experience. By my estimations, if I can be tricked even at my most intense doubts, then the trick is so convincing it is practically the real thing, and a practical intelligence is intelligence.

 

Let us assume that we can define a metric, a Resolution of Autonomy. At the bottom of this scale, 0, is a rock. At the top, 100, there is a human being. At any given moment a Tulpa is somewhere on this sliding scale, and the goal of the entire process is to slide the Tulpa from the bottom to the top.

 

The problem is this number is forever unknown, the metric is fundamentally unknowable even if the mind exists in OUR mind (https://en.wikipedia.org/wiki/Problem_of_other_minds). This leads to the existence of another metric, the Perceived Resolution. The top and bottom of this scale are identical to the other, except this time what is measured is what the observer estimates the subject's Resolution of Autonomy to be at.

 

A child talking to their stuffed toy has a P.A of 100, but the R.A is 0. An NPC in a videogame could be somewhere around an R.A of 6, but the brain plays tricks and gives it a P.A of 8 because of their human like animations and voice lines.

 

But I offer this thought: In the creation of a Tulpa, if the P.A exceeds the R.A, it is indistinguishable to the observer from them being equal.

 

If the P.A raises, the mind will stop restricting the natural "imaginary friend" control of the Tulpa, because such restraints are only necessary during early stages. This Tulpa will now be a mundane imagined character, but the host will believe a process has made them autonomous. This is extremely troubling, because the conclusion we have to come to is that we cannot truly trust the sentience of a Tulpa. I do not want this to be the true, and want to find a logical way out. (That is to say, we cannot trust it ontologically. You should trust your Tulpa in the real world! Don't get solipsistic on me, dude!)

 

Herein lies the rub ,though - Is a Tulpa with a P.A of 100 and an R.A of 100 any different than one of 100-0? The natural answer is "Duh". I ask then what the fundamental difference between them is. The host fully believes they have a sentient Tulpa in both cases, in fact in both cases their mental states are indistinguishable. In both states the host gets results from their Tulpa which satisfies their requirements for autonomy, both of them believe the Tulpa is a foreign entity, although only one of them is right.

 

Are both of them Tulpa? Is there a tick on the R.A scale that defines the lowest bar a Tulpa has to reach before graduating from a bundle of thoughts?

 

Is a Tulpa a phenomenon that's reality is marked not by ontological existence, but practical function to the host?

 

Certainly, all of us here believe there is something more at play. But how to we define that? Can we even produce a reliable definition of what a Tulpa is?

 

Everyone marks this feeling where they can just kind of...know. They have no control, this is a foreign entity. How do we actually break down and examine this moment beyond a feeling? What are the mechanisms at play?

 

We know obviously the overall thing a Tulpa SHOULD be: An imagined being that has been made to autonomously function to a point of achieved individual self.

 

 

This is, at the end of the day, a philosophical challenge. This of course doesn't mean your Tulpa doesn't exist and you should question everything. Rather, this is a challenge of definitions and states that I wonder if we can overcome.

 

 

 

My particular method of forcing is essentially trying to keep the P.A matched with the R.A as rigorously as possible, to only let anything happen if something demonstrably exceeds my P.A, thus forcing adaptation that grows the R.A

 

How this actually WORKS though is probably the biggest mystery in the process. I hypothesized earlier that intrusive thoughts under scrutiny will have to adapt to continue, but why are intrusive thoughts adapting? What mechanism is making them adapt? Intrusive thoughts don't mutate like DNA, surely.

 

Perhaps there is a stage of Shared Autonomy, where we lend a Tulpa little bit of our own autonomy to them and guide the process. This is done instantaneously so we cannot comprehend it, and we harmlessly interpret it as their own autonomy, and what you work towards to get the first steps. Then perhaps as the Active Model is practiced enough, it can gain singular Owned Autonomy. This is definitely the hardest thing to model by my estimations.

 

But I think a good definition for Owned Autonomy, the final goal, is the moment where the Tulpa's own mechanisms are making predictions and modifying itself faster than you can possibly consciously intervene.

 

(PSSSSSSSST! IF YOU MADE IT THIS FAR AND ARE AN EXPERIENCED TULPA AND ARE LIKE "I CAN THINK OF LIKE 5 REASONS THIS DOESN'T MAKE SENSE" YOU SHOULD LIKE, TOTALLY TELL ME SO I CAN LEARN AND GET A BETTER MODEL :])

 

 

 

Well, if you read that, I guess we both had nothing to do today.

 

Oh! I JUST remembered I actually have one of these Tulpa things and I'm supposed to be like, creating it. Let's do that!

 

Today is a repeat of yesterday, lock in on the visualization and personality forcing with both closed and open eyed techniques. 

 

---POST SESSION---

 

     Spent about 1-2ish hours actively forcing. Focused on visualization, and sense things were moving along smoothly I focused on looking at my Tulpa from all angles and getting specifics correct. Tried something new with personality forcing: Went into my memories of the past few days and puppetted my Tulpa there with me, simulating what they would've done. After a bit more puppetting practice, I went outside and walked in the rainy night doing narration the whole time. Narration felt massive in increasing permanence, I can "feel" my Tulpa in my head right now. So far, I have successfully maintained a constant feeling that my work is stacking up and I'm getting a better idea of both what my Tulpa should look like, and what they should act like.

 

    Keeping my Tulpa insulated from my own existence is going well enough. I'm careful to choose what I am narrating wisely and am trying to mentally filter out any data that hints at my existence, but we will see how powerful these memory "suggestions" are, and if my Tulpa can just access the raw memories. Exciting stuff!

 

    Narration will be crucial going forward, so I'm going to try to keep it up daily no matter what.

 

That's all for now!!!

 

[Mai_x_v3]

3 hours ago, Biscuit said:

Thank you for your kind words!!! All is going perfectly swell now luckily! Thank you for caring enough to read, and it's always a pleasure to see a fellow V3 believer out in the wild

You're welcome and of course! And yeah I don't really feel comfortable with doubles but thank you! 🩷💕 I'm not sure if you meant the game though or something else because my tag is a shorter end of the last two aberration of a game I've been obsessed with since 2017. Don't take that the wrong way either, it's just I've always been uncomfortable when others starting talking about that particular special interest to me, especially when it's out of the blue too, it's nothing against anyone either. It just makes me feel uncomfortable talking about it, ya know? Also, amazing log this time good luck at future forcing! You got this!

 

It is cool to see other fans though, I think once my tulpas become fully vocal these feelings would subsided since I'm also trying to heal from the uncomfortableness, if that makes sense. But I won't bother you with getting too personal. I don't want to clog up your log that'll be impolite of me. My DMs are always open though.

Edited October 11, 2025 by Mai_x_v3

[Biscuit]

5 hours ago, Mai_x_v3 said:

You're welcome and of course! And yeah I don't really feel comfortable with doubles but thank you! 🩷💕 I'm not sure if you meant the game though or something else because my tag is a shorter end of the last two aberration of a game I've been obsessed with since 2017. Don't take that the wrong way either, it's just I've always been uncomfortable when others starting talking about that particular special interest to me, especially when it's out of the blue too, it's nothing against anyone either. It just makes me feel uncomfortable talking about it, ya know? Also, amazing log this time good luck at future forcing! You got this!

 

It is cool to see other fans though, I think once my tulpas become fully vocal these feelings would subsided since I'm also trying to heal from the uncomfortableness, if that makes sense. But I won't bother you with getting too personal. I don't want to clog up your log that'll be impolite of me. My DMs are always open though.

Gotcha gotcha, I understand! Thanks for reading and thank you for the encouragement!!!!

[Mai_x_v3]

33 minutes ago, Biscuit said:

Gotcha gotcha, I understand! Thanks for reading and thank you for the encouragement!!!!

Thank you for understanding! And you're welcome! And ofc I'm looking forward to so many of your amazing updates I know you can do it! You're so smart and intelligent and throughall! 💕🩷

[Biscuit]

Weeks 1-2

Days 3-8

 

It's over. The organization has circled in, the agents pounding on my door. There is little time left, they will soon reach the cipher locked in my brain. This progress report will self destruct in the next 30 seconds. Godspeed.

 

I'm kidding (you better hope I am)

 

We've reached week 2, so I found it proper to do a little check in even if I haven't made any groundbreaking progress. Things have been very by the book so I thought it might be nice to shed a light on the smaller things so this isn't such a dry read of a PR. 

Overview

     Things have mostly been going extremely well on the front of visualization. I started with a form that had a lot of references, and sticking to those I would say I am nearing a complete image of my Tulpa. However, things are taking strange turns. My mind seems to be unglueing the form that I had set up and attempting to sway me ever-so-slightly towards another form (specifically, a form from a failed attempt many years ago). This is sort of baffling, as I have no idea where these thoughts are coming from. Perhaps firing away at familiar behavior is exciting old memories and bubbling them up? It seems too early by my estimations to be some kind of intelligent divergence, more likely automatic thoughts. Still, the "weight" of my Tulpa seems to naturally be shifting and almost easier to bring up with this new form. My mind is forming tokens without my consent, which is not unlike how actual memories are made. I'm still deciding if it is best to follow the path of least resistance and accept this new form, or stick to my guns. I had made the original form black and white to see how it became colored, and this would ruin that aspect, but it may make the process more natural. The name I chose is also being overwritten in my automatic thoughts with another name I was considering, so I think I'll go with that one as I want to err on the side of spontaneity to encourage growth.

     Personality forcing is a real slog. I haven't been focused enough to do long engagement puppeting exercises, and keep starting late into the night when I'm already tired. This will not do in the future so I need to start getting on it in the mornings. I have the time tested 50 metric ton list of personality traits, and even some musing on what those entail. While I won't break down my exact techniques here (I'll save that for when I have a breakthrough so we can see what worked and what didn't), it is definitely hard to do when I'm tired. (A lot of memorization, repeating off of lists, and ideally puppeting. You get the picture - pretty classic stuff)

     The black box experiment is all-encompassing and has markedly set this experience apart from my other attempts. Nearly everything has to be rethought out and planned to insulate my existence from my Tulpa. I become a disembodied voice in my mind's eye, anything to give my psyche an obvious clue that I don't want my Tulpa to notice me. It can be guessed that memories from this stage are retroactively given when autonomy is reached, so I am hoping that via making a conscious effort to remind myself that I am separating my identity from this process that none of that will be tainted when that process begins.

Narration is going tediously well. I keep forgetting to narrate and then get quick and valuable bursts in, but I really wish I had more presence with it, so that will be a goal for this week.

                                                                                                                                                                                                         Overall I believe there are two places I am failing: Empathetic Understanding and Anticipation.  

   By Empathetic Understanding, I mean I actually don't have a very good feel for the person my Tulpa is supposed to be. I loathe to admit that my personality forcing has probably been much too mild and the model I have is far too clinical and thin. I need to work more on actually building an understanding of a person, which shouldn't be too hard if I actually focus on it and let myself actually feel things about my Tulpa instead of being so cut and dry about it.

   By Anticipation I mean I am not casting enough moments of expectation to let my mind fill them. I have another article I wrote about this below spoilered if you are interested.

Spoiler

The natural, nearly mutative growth of Tulpas still seems to be a mystery. In most cases it is hand-waved off as the "subconscious" while not actually determining an observable system of subconscious thought, making it a useless magic word explanation. 

While this post will not make a model for Tulpa Autonomy in general, I want to make a case for a system of adaptation that relies on anticipation. 

 

Largely as a community there is a consensus that intent has a bearing on the process. "Assume sentience from the start" has been mostly seen as a cognitive framing tool to self-fulfill-your-prophecy to sentience (i.e beget behavior that reinforces and leads to sentience), and there is a lot of thought that a doubtful mindset tends to lead to slower development because you cannot interpret any of your brain's attempts as successful. 

 

I believe this idea of "intent" marries well with a general idea of anticipation. At the heart of this idea is this proposition: A Tulpa is far more likely to do something if they are anticipated to act. It could, controversially, even be extended to say that Tulpas literally *cannot* act if action is not anticipated in some way (more on this admittedly radical idea later). 

 

The brain has extremely strong predictive mechanisms, as evidenced by the Predictive Coding model of cognition. I'm not asking that you subscribe to it, but rather want to point out that there is seemingly undeniable evidence to at least our profound skill at modeling and predicting behavior. Tulpas exist in purely a cognitive medium, and with deduction it is possible to come to the conclusion that they use very similar facilities to predictive modeling*.

 

We can consider anticipation a period of time where we are awaiting a response we believe may come. When we knock on a door, there is a period of anticipation as to what will happen. For many the mind might run wild imagining situations and preparing verbal responses, or we may instead worry about how we look (spurred on by imagining another person opening the door and seeing us), the list is endless but the period of anticipation is pretty undeniable.

 

Anticipation seems to activate predictive mechanisms in the brain related to what is happening, and I am arguing that this Anticipatory Excitement of the brain is a tool used in Tulpa automation.

 

There is a common (or at least was common, not sure anymore!) idea that if you are having trouble with vocality that you aren't "listening" for your Tulpa, and this mindset actually worked in many cases of reputable hosts. These testimonies can be explained with Anticipatory Excitement, because their brain was "listening" (anticipating a response), they began to excite parts of their brain that predicted and imagined responses. With genuine cognitive mechanisms simulating what a Tulpa might say it becomes easier for the active model of the Tulpa to choose one/synthesize one, whichever of these ascribes to your belief.

 

In this case, puppeting and parroting become artifacts of Anticipatory Excitement. Your brain is imagining what they will say/do against your will and applying it to the imagined space in your head, even though these are separate entirely from the Tulpa model. It stands to reason to that learned anticipation can be dangerous to the process too: If you train yourself to only anticipate silence then perhaps silence is all you can ever receive.

 

Therefore we can assume that when making a Tulpa it is vital to not only foster these fields of anticipation but to also make sure you are not reinforcing silence. This does beg a question, if we don't stifle these predicted responses, how can we ever determine  unrelated predictions vs. Tulpa thoughts if they use the same system? I think this requires a stronger modelling of Tulpa Autonomy in general, but I don't think it defeats the purpose of what this post is saying. 

 

These thoughts have existed in other words throughout the community, and I wanted to bring them together in a solidified idea, and also consider it using predictive processes. I think there is much more to say in how our predictive mechanisms are the likely culprits as to why this is all possible, but that is far beyond the scope of this post. 

To extend a more radical idea just to get thoughts out there, perhaps anticipation is literally required for a Tulpa to function. Maybe in high-complexity Tulpas that are fully autonomous the brain learns to instantaneously throw out a moment of anticipation when certain stimuli associated with them are experienced or when the host thinks of the Tulpa? Ultimately I find this to be bending the definition of anticipation too much, and this slides more into the autonomously using predictive mechanisms, but I wanted to put the idea out there nonetheless. 

 

 

 

 

 

 

*My reasoning behind this: The same mechanism used to simulate how someone we know would react to a situation likely drives a Tulpas actions. Tulpas conceptually are built on our own predictive understanding of how consciousness will act based on a wealth of experience. Their form is also tied to how we understand materials react in the physical world based on a predictive model. Although I will admit its not near bulletproof, I think it is reasonable to assume Tulpas and predictive modeling are closely related in that Tulpas hijack some of that functionality, or more precisely we use those tools in novel ways to generate Tulpas.  

Essentially: The black box environment has removed the call and response timings of normal conversation. Starting today, I have tried to set distinct moments where I make something happen and wait for a response, welcoming floods of thoughts but just letting them pass over. I think this will be very good for the future.

But other than that, not too much to report! Things are going steady, I want to put more time into forcing as I'm just getting pockets here and there wherever I can. I definitely have the time, I just need to get more comfortable just going for it. 

Bye for now!

Edited October 19, 2025 by Biscuit