Is it possible to forget a information?

I stored a xyz information yesterday, now I can retrieve the information by giving proper input. Is it possible to forget a information if I don’t give proper input?
If yes or no ! Whats your explanation about it?

It will be eventually forgotten (if implemented close to brain mechanisms) in the absence of the same/similar input. Think of weakening synapses, which will become dormant/erased after non-use. Use it or lose it principle of neuroplasticity. Time required for complete forgetting is very variable. Short term memory is forgotten very quickly if not repeated, emotionally significant/strong synapses may never be really forgotten.

1 Like

Introspection being what it is …

I know that I can (under the right conditions) remember things that I would have claimed that I have well and truly forgotten.

Some background - I often keep a travel diary when I am traveling for work. These are usually fairly detailed with personal observations. On trips from 20 years ago, to cities where I can’t even remember the name of, when I pull out the written record and start reading an amazing (to me) amount of detail comes back to me. It was all in my head somewhere, waiting for the right key to recall it.

So the information was forgotten; without the probe I doubt that I would ever have accessed it again.
It is highly likely that everyone has “forgotten” things that could be accessed under the right conditions.

This leads to more basic questions; what is perceiving and remembering?

Now I am treading on wild conjecture…

It is also highly likely that the “forgotten” information becomes part of the vast cache of information that makes up your personality; one of the numberless factoids that join together whenever the active process that we name as perception is functioning.

When you experience something your brain responds by assembling the closest bits of memory to match the perception and build an internal model on the fly. If you have experienced something before the memory fragments are joined to make up this perception. Over time these bits are combined and generalized. The original memory may still be present but now it does duty in a more general sense.


I think the brain never tries to purposefully delete anything, I think you can replace one association with another but in reality you would have two attractors where the most recent one is easier to reach.

you can probably “delete” information by storing random data in the memory system until it gets saturated and starts forgetting older memories but then you’d just be destroying the memory, not just selectively deleting one association.

maybe you can get a similar effect to a “deletion” by inserting a repeated association in the system like:

A — > B
suppose we have a system that associates “A” to “B”, essentially storing “B” onto memory adress “A”

A — > C
if we try to store a new association:

result = (B + C)
when we retrieve address A we get a superposition of B and C

but if we keep storing C onto A over and over:
A — > C
A — > C
A — > C

result = (0.25B + 0.75C)

we get a vanishingly small contribution from B until its negligible.


The information that wasn’t retrieved, can we assume that the information is forgotten.
Eg. Remaining 0.75 percent of B information.

Is it possible that information was never forgotten in the brain? Can we assume that?
And also can we assume the information was only forgotten bcoz you never saved it or information never got into the mind?

So anything that is stored, is never forgotten.
So the things that is forgotten, is never stored in the first place.

well, it depends, if the information is stored as a very large holographic distributed encoding i.e an SDR then 0.01% of bits probably contain enough information to reconstruct the whole symbol, so no information is lost until it reaches nearly zero.


I’ll try to make a a quick conservative estimate, please don’t take this too seriously.

lets assume a neuron in cortex can choose to connect to any other neuron in a pool of 100k potential connections.
so we need about 17bits to describe a single binary connection.

and lets assume a neuron can have about 10000 connected synapses at any given time, so a single neuron could store 17bits * 10000 = 21.25 kilobytes

according to google the brain contains about 86billion neurons but lets assume only 10billion can be used to store memories;
so thats 10billion * 21.25 kilobytes = 212.5 terabytes

the optic nerve contain about one million fibers, and humans can distinguish about 30 levels of luminosity so we’d need at most 5bits * one million fibers * 2 eyes = 1.25 megabytes

according to google, the average compression ratio for a jpeg image is 10:1 so lets assume the primary visual cortex can do something similar with its gabor filters, that reduces 1.25Mb down to 125 kilobytes.

if we assume the brain will store one visual snapshot every 2 seconds on average.
125 kilobytes * 30 snapshots per minute * 60 minutes * 16 awake hours * 365 days * 100 years = 131.4 terabytes

So… maybe?


That was some insane explanation… I like the way you described it… But we really forget the stored informations? Even if we assume stored information can’t be forgotten, it will be wrong? Right! Time is the destroyer of our information? How do we solve this problem?

What destroys a stored information?

I think the information inevitably deteriorates with time as a side effect of the stochastic nature of biological networks, synapses move here and there, bits get randomly flipped, but the brain is build to tolerate those defects really well by having lots of redundancy, so a memory needs to deteriorate to the point of being indistinguishable from noise. which may take a long time but will happen eventually.

but I’d say thats a limitation rather than a feature.

1 Like