The
Mandela
Effect
is
an
internet
phenomenon
describing
shared
and
consistent
false
memories
for
specific
icons
in
popular
culture.
Visual
(VME)
a
to
visual
(e.g.,
the
Monopoly
Man
falsely
remembered
with
monocle)
has
not
yet
been
empirically
quantified
or
tested.
In
Experiment
1
(N=100),
we
demonstrate
that
certain
images
from
iconography
elicit
consistent,
memories.
2
(N=60),
using
eye-tracking-like
methods,
find
no
attentional
differences
drive
this
phenomenon.
There
clear
difference
natural
experience
of
these
(Experiment
3),
VME-errors
also
occur
spontaneously
during
recall
4;
N=50).
These
results
there
are
which
people
consistently
make
same
memory
error,
despite
majority
being
canonical
image.
Science Advances,
Journal Year:
2023,
Volume and Issue:
9(17)
Published: April 26, 2023
What
makes
certain
images
more
memorable
than
others?
While
much
of
memory
research
has
focused
on
participant
effects,
recent
studies
using
a
stimulus-centric
perspective
have
sparked
debate
the
determinants
memory,
including
roles
semantic
and
visual
features
whether
most
prototypical
or
atypical
items
are
best
remembered.
Prior
typically
relied
constrained
stimulus
sets,
limiting
generalized
view
underlying
what
we
remember.
Here,
collected
1
million
ratings
for
naturalistic
dataset
26,107
object
designed
to
comprehensively
sample
concrete
objects.
We
establish
model
that
is
predictive
image
memorability
examined
could
be
accounted
by
typicality
find
exert
stronger
influence
perceptual
remember
relationship
between
complex
simple
positive
negative
association
alone.
Cognition,
Journal Year:
2024,
Volume and Issue:
248, P. 105810 - 105810
Published: May 11, 2024
Human
observers
often
exhibit
remarkable
consistency
in
remembering
specific
visual
details,
such
as
certain
face
images.
This
phenomenon
is
commonly
attributed
to
memorability,
a
collection
of
stimulus
attributes
that
enhance
the
long-term
retention
information.
However,
exact
contributions
memorability
memory
formation
remain
elusive
these
effects
could
emerge
anywhere
from
early
perceptual
encoding
post-perceptual
consolidation
processes.
To
clarify
this,
we
tested
three
key
predictions
hypothesis
facilitates
supports
short-term
(VSTM)
and
(VLTM).
First,
examined
whether
benefits
VSTM
manifest
early,
even
within
constraints
brief
presentation
(100-200
ms;
Experiment
1).
We
achieved
this
by
manipulating
duration
change
detection
task
using
images
with
high-
or
low-memorability
while
ensuring
they
were
equally
familiar
participants.
Second,
assessed
benefit
increases
likelihood
retention,
post-stimulus
masking
designed
interrupt
processes
(Experiment
2).
Last,
investigated
durability
intervals
seconds
24
h
3).
Across
experiments,
our
data
suggest
has
an
impact
on
formation,
persisting
across
variable
predicting
subsequent
VLTM
overnight.
Combined,
findings
highlight
enhances
100-200
ms
following
onset,
resulting
robust
traces
resistant
interruption
forgetting.
Sensors,
Journal Year:
2025,
Volume and Issue:
25(6), P. 1661 - 1661
Published: March 7, 2025
The
accurate
modelling
of
video
memorability,
or
the
intrinsic
properties
that
render
a
piece
audiovisual
content
more
likely
to
be
remembered,
will
facilitate
development
automatic
systems
are
efficient
in
retrieving,
classifying
and
generating
impactful
media.
Recent
studies
have
indicated
strong
correlation
between
visual
semantics
its
memorability.
This
underscores
importance
developing
advanced
comprehension
abilities
enhance
model
performance.
It
has
been
demonstrated
Large
Vision–Language
Models
(LVLMs)
demonstrate
exceptional
proficiency
generalist,
high-level
semantic
images
video,
due
their
extensive
multimodal
pre-training
on
vast
scale.
work
makes
use
generalist
knowledge
LVLMs
explores
adaptation
techniques
with
view
utilising
them
as
memorability
predictors.
In
particular,
Quantized
Low-Rank
Adaptation
(QLoRA)
technique
is
employed
fine-tune
Qwen-VL
memorability-related
data
extracted
from
Memento10k
dataset.
light
existing
research,
we
propose
particular
methodology
transforms
language
score
regressor.
Furthermore,
consider
influence
selecting
appropriate
LoRA
hyperparameters,
design
aspect
insufficiently
studied.
We
validate
rank
alpha
hyperparameters
using
5-Fold
Cross-Validation
evaluate
our
best
configuration
official
testing
portion
dataset,
obtaining
state-of-the-art
Spearman
Rank
Correlation
Coefficient
(SRCC)
0.744.
Consequently,
this
represents
significant
advancement
through
understanding.
Psychological Science,
Journal Year:
2022,
Volume and Issue:
33(12), P. 1971 - 1988
Published: Oct. 11, 2022
The
Mandela
effect
is
an
Internet
phenomenon
describing
shared
and
consistent
false
memories
for
specific
icons
in
popular
culture.
visual
a
to
(e.g.,
the
Monopoly
Man
falsely
remembered
as
having
monocle)
has
not
yet
been
empirically
quantified
or
tested.
In
Experiment
1
(N
=
100
adults),
we
demonstrated
that
certain
images
from
iconography
elicit
consistent,
memories.
2
60
using
eye-tracking-like
methods,
found
no
attentional
differences
drive
this
phenomenon.
There
clear
difference
natural
experience
of
these
(Experiment
3),
errors
also
occur
spontaneously
during
recall
4;
N
50
adults).
These
results
demonstrate
there
are
which
people
consistently
make
same
false-memory
error,
despite
majority
being
canonical
image.
bioRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2022,
Volume and Issue:
unknown
Published: April 30, 2022
ABSTRACT
What
makes
certain
images
more
memorable
than
others?
While
much
of
memory
research
has
focused
on
participant
effects,
recent
studies
employing
a
stimulus-centric
perspective
have
sparked
debate
the
determinants
memory,
including
roles
semantic
and
visual
features
whether
most
prototypical
or
atypical
items
are
best
remembered.
Prior
typically
relied
constrained
stimulus
sets,
limiting
generalized
view
underlying
what
we
remember.
Here,
collected
1+
million
ratings
for
naturalistic
dataset
26,107
object
designed
to
comprehensively
sample
concrete
objects.
We
establish
model
that
is
predictive
image
memorability
examined
could
be
accounted
by
typicality
find
exert
stronger
influence
perceptual
remember
relationship
between
complex
simple
positive
negative
association
alone.
TEASER
Semantic
versus
heavily
remember,
cannot
reduced
typicality.
Neuropsychologia,
Journal Year:
2021,
Volume and Issue:
160, P. 107976 - 107976
Published: July 24, 2021
Endel
Tulving's
episodic
memory
framework
emphasizes
the
multifaceted
re-experiencing
of
personal
events.
Indeed,
decades
research
focused
on
experiential
nature
memories,
usually
treating
recent
as
a
coherent
quality.
However,
insights
into
functional
architecture
medial
temporal
lobe
show
that
different
types
mnemonic
information
are
segregated
distinct
neural
pathways
in
brain
circuits
empirically
associated
with
memory.
Moreover,
memories
do
not
fade
whole
under
conditions
progressive
neurodegeneration
these
circuits,
notably
Alzheimer's
disease.
Instead,
certain
content
seem
particularly
vulnerable
from
moment
their
encoding
while
other
can
remain
memorable
consistently
across
individuals
and
contexts.
We
propose
observations
related
to
content-specific
consequently
impairment
at
stages
neurodegeneration.
To
develop
inspirational
legacy
further
advance
our
understanding
how
function
is
affected
by
neurodegenerative
such
disease,
we
postulate
it
compelling
focus
representational
memories.
Learning & Memory,
Journal Year:
2020,
Volume and Issue:
27(12), P. 503 - 509
Published: Nov. 16, 2020
The
features
of
an
image
can
be
represented
at
multiple
levels—from
its
low-level
visual
properties
to
high-level
meaning.
What
drives
some
images
memorable
while
others
are
forgettable?
We
address
this
question
across
two
behavioral
experiments.
In
the
first,
different
layers
a
convolutional
neural
network
(CNN),
which
represent
progressively
higher
levels
features,
were
used
select
that
would
shown
100
participants
through
form
prospective
assignment.
Here,
discriminability/similarity
with
others,
according
CNN
dictated
presented
groups,
who
made
simple
indoor
versus
outdoor
judgment
for
each
scene.
found
remember
more
scene
selected
based
on
their
discriminability
or
similarity.
A
second
experiment
replicated
these
results
in
independent
sample
50
participants,
order
postencoding
tasks.
Together,
experiments
provide
evidence
both
and
similarity,
levels,
predict
memorability.
Journal of Cognitive Neuroscience,
Journal Year:
2023,
Volume and Issue:
35(4), P. 603 - 627
Published: Jan. 1, 2023
Abstract
Despite
its
unlimited
capacity,
not
all
visual
information
we
encounter
is
encoded
into
long-term
memory.
Traditionally,
variability
in
encoding
success
has
been
ascribed
to
the
types
and
efficacy
of
an
individual's
cognitive
processes
during
encoding.
Accordingly,
past
studies
have
identified
several
neural
correlates
success,
namely,
frontal
positivity,
occipital
alpha
amplitude,
theta
by
contrasting
electrophysiological
signals
recorded
successful
failed
(i.e.,
subsequent
memory).
However,
recent
research
demonstrated
individuals
remember
forget
consistent
sets
stimuli,
thereby
elucidating
stimulus-intrinsic
factors
memorability)
that
determine
ease
memory
independent
individual-specific
processes.
The
existence
memorability
raises
possibility
canonical
EEG
may
reflect
rather
than
To
test
this,
while
participants
600
images
real-world
objects
assessed
unique
contribution
on
each
correlate.
Here,
found
amplitude
were
only
influenced
whereas
positivity
was
success.
Overall,
our
results
offer
novel
interpretations
demonstrating
a
dissociable
impact