NeuroImage,
Journal Year:
2024,
Volume and Issue:
291, P. 120602 - 120602
Published: April 4, 2024
Working
memory
(WM)
describes
the
dynamic
process
of
maintenance
and
manipulation
information
over
a
certain
time
delay.
Neuronally,
WM
recruits
distributed
network
cortical
regions
like
visual
dorsolateral
prefrontal
cortex
as
well
subcortical
hippocampus.
How
input
dynamics
subsequent
neural
impact
remains
unclear
though.
To
answer
this
question,
we
combined
analysis
behavioral
capacity
with
measuring
through
task-related
power
spectrum
changes,
e.g.,
median
frequency
(MF)
in
functional
magnetic
resonance
imaging
(fMRI).
We
show
that
processing
dynamics,
task
structure's
specific
timescale,
leads
to
changes
unimodal
cortex's
corresponding
timescale
which
also
relates
working
capacity.
While
more
transmodal
hippocampus
its
balance
across
multiple
timescales
or
frequencies.
In
conclusion,
here
relevance
both
different
for
uni
-
subject's
performance.
The
ability
to
sustain
internal
representations
of
the
sensory
environment
beyond
immediate
perception
is
a
fundamental
requirement
cognitive
processing.
In
recent
years,
debates
regarding
capacity
and
fidelity
working
memory
(WM)
system
have
driven
significant
advances
in
our
understanding
nature
these
representations.
particular,
there
growing
recognition
that
WM
are
not
merely
imperfect
copies
perceived
object
or
event,
as
new
experimental
tools
revealed
observers
possess
richer
information
about
uncertainty
their
memories,
take
advantage
environmental
regularities
use
limited
resources
optimally.
Meanwhile,
computational
models
visual
formulated
at
different
levels
implementation
converged
on
common
principles
relating
variability
uncertainty.
Here
we
review
research
human
from
perspective
latest
developments
identifying
neural
mechanisms
support
it.
Current Biology,
Journal Year:
2022,
Volume and Issue:
32(6), P. 1275 - 1284.e4
Published: Feb. 14, 2022
Arithmetic
is
a
cornerstone
of
scientifically
and
technologically
advanced
human
culture,
but
its
neuronal
mechanisms
are
poorly
understood.
Calculating
with
numbers
requires
temporary
maintenance
manipulation
numerical
information
according
to
arithmetic
rules.
We
explored
the
brain
involved
in
simple
operations
by
recording
single-neuron
activity
from
medial
temporal
lobe
subjects
performing
additions
subtractions.
found
abstract
notation-independent
codes
for
addition
subtraction
populations.
The
different
areas
differed
drastically.
Decoders
applied
time-resolved
recordings
demonstrate
static
code
hippocampus
based
on
persistently
rule-selective
neurons,
contrast
dynamic
parahippocampal
cortex
originating
neurons
carrying
rapidly
changing
rule
information.
implementation
suggests
cognitive
functions
regions
arithmetic.
Cognition,
Journal Year:
2022,
Volume and Issue:
231, P. 105330 - 105330
Published: Nov. 24, 2022
The
present
experiments
support
two
conclusions
about
the
capacity
limit
of
working
memory
(WM).
First,
they
provide
evidence
for
Binding
Hypothesis,
WM
is
limited
by
interference
between
bindings
but
not
items.
Second,
show
that
episodic
LTM
contributes
substantially
to
binding
when
stretched
larger
set
sizes.
We
tested
immediate
sets
word-picture
pairs.
With
increasing
size,
declined
more
precipitously
than
items,
as
predicted
from
hypothesis.
Yet,
at
higher
sizes
performance
was
stable
expected
a
memory,
suggesting
contribution
long-term
(LTM)
circumvent
limit.
In
hypothesis,
we
double
dissociation
contributions
and
memory:
Performance
3
specifically
affected
proactive
–
were
immune
influences
distractor-filled
delay.
contrast,
size
2
unaffected
harmed
Nature Human Behaviour,
Journal Year:
2023,
Volume and Issue:
8(2), P. 336 - 348
Published: Dec. 18, 2023
Stimulus-dependent
eye
movements
have
been
recognized
as
a
potential
confound
in
decoding
visual
working
memory
information
from
neural
signals.
Here
we
combined
eye-tracking
with
representational
geometry
analyses
to
uncover
the
miniature
gaze
patterns
while
participants
(n
=
41)
were
cued
maintain
object
orientations.
Although
discouraged
breaking
fixation
by
means
of
real-time
feedback,
small
shifts
(<1°)
robustly
encoded
to-be-maintained
stimulus
orientation,
evidence
for
encoding
two
sequentially
presented
orientations
at
same
time.
The
orientation
on
presentation
was
object-specific,
but
it
changed
more
object-independent
format
during
maintenance,
particularly
when
attention
had
temporarily
withdrawn
memorandum.
Finally,
categorical
reporting
biases
increased
after
unattended
storage,
indications
biased
geometries
already
emerging
maintenance
periods
before
behavioural
reporting.
These
findings
disclose
wealth
visuospatial
and
indicate
systematic
changes
contents
unattended.
Journal of Cognitive Neuroscience,
Journal Year:
2022,
Volume and Issue:
34(12), P. 2360 - 2374
Published: Jan. 1, 2022
Abstract
Although
storage
in
working
memory
(WM)
can
be
tracked
via
measurements
of
ongoing
neural
activity,
past
work
has
shown
that
observers
maintain
access
to
information
despite
temporary
interruptions
those
patterns.
This
observation
been
regarded
as
evidence
for
a
neurally
silent
form
WM
storage.
Alternatively,
however,
unattended
could
retrieved
from
episodic
long-term
(eLTM)
rather
than
being
maintained
during
the
activity-silent
period.
Here,
we
tested
between
these
possibilities
by
examining
whether
performance
showed
proactive
interference
(PI)—a
hallmark
retrieval
eLTM—following
such
interruptions.
Participants
remembered
colors
(Experiments
1–3)
or
locations
(Experiment
4)
serially
presented
objects.
We
found
PI
set
sizes
larger
4,
but
not
smaller
sizes,
suggesting
eLTM
may
have
supported
when
capacity
was
exceeded.
Critically,
with
small
remained
resistant
PI,
even
following
prolonged
challenging
distractor
task.
Thus,
PI-resistant
memories
were
across
likely
storage-related
an
empirical
pattern
implies
WM.
NeuroImage,
Journal Year:
2023,
Volume and Issue:
272, P. 120055 - 120055
Published: March 30, 2023
Evolving
behavioral
goals
require
the
existence
of
selection
mechanisms
that
prioritize
task-relevant
working
memory
(WM)
content
for
action.
Selecting
an
item
stored
in
WM
is
known
to
blunt
and/or
reverse
information
loss
stimulus-specific
representations
reconstructed
from
human
brain
activity,
but
extant
studies
have
focused
on
all-or-none
circumstances
allow
or
disallow
agent
select
one
several
items
WM.
Conversely,
suggest
humans
can
flexibly
assign
different
levels
priority
WM,
how
doing
so
influences
neural
unclear.
One
possibility
assigning
quality
those
representations,
resulting
more
robust
high-
vs.
low-priority
content.
A
second
–
and
non-exclusive
asymmetries
influence
rapidly
be
selected
reported.
We
tested
these
possibilities
two
experiments
by
decoding
EEG
recordings
obtained
while
volunteers
performed
a
retrospectively
cued
task.
Probabilistic
changes
relevance
remembered
had
no
effect
our
ability
decode
it
signals;
instead,
influenced
latency
at
which
above-chance
performance
was
reached.
Thus,
results
indicate
probabilistic
ease
with
memories
independently
their
strength.
Psychological Review,
Journal Year:
2023,
Volume and Issue:
131(2), P. 321 - 348
Published: June 15, 2023
Whether
working
memory
reflects
a
thresholded
recollection
process
whereby
only
limited
number
of
items
are
maintained
in
memory,
or
signal
detection
which
each
studied
item
is
increased
familiarity
strength,
topic
considerable
debate.
A
review
visual
studies
that
have
examined
receiver
operating
characteristics
(ROCs)
across
broad
set
materials
and
test
conditions
indicates
Cerebral Cortex,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 5, 2025
Recent
studies
have
proposed
that
visual
information
in
working
memory
(WM)
can
be
maintained
an
activity-silent
state
and
reactivated
by
task-irrelevant
high-contrast
impulses
("ping").
Although
pinging
the
brain
has
become
a
popular
tool
for
exploring
WM,
its
underlying
mechanisms
remain
unclear.
In
current
study,
we
directly
compared
neural
reactivation
effects
behavioral
consequences
of
spatial-nonmatching
spatial-matching
pings
to
distinguish
noise-reduction
target-interaction
hypotheses
brain.
Initially,
electroencephalogram
our
decoding
results
showed
WM
transiently
without
changing
original
representations
or
recall
performance.
Conversely,
more
durably
further
reorganized
decreasing
representations'
dynamics.
Notably,
only
strength
correlated
with
performance
was
modulated
location
memorized
items,
occurring
when
both
items
were
presented
horizontally.
Consistently,
follow-up
found
spatial-matching,
horizontal
impaired
no
ping.
Together,
demonstrated
two
distinct
brain,
highlighting
critical
role
ping's
context
(i.e.
spatial
information)
reactivating
reorganizing
WM.
Frontiers in Psychology,
Journal Year:
2025,
Volume and Issue:
16
Published: April 9, 2025
To
support
goal-directed
behavior,
working
memory
(WM)
must
flexibly
access
relevant
information.
While
the
mechanisms
underlying
single-item
WM
are
comparatively
well-studied,
less
is
known
about
principles
governing
multi-item
access.
Some
studies
have
suggested
that
dual-item
retrieval
can
be
as
efficient
access,
but
it
remains
unclear
whether
this
reflects
reduced
inhibitory
demands
or
truly
parallel,
cost-free
retrieval.
In
Experiment
1,
we
manipulated
number
of
vs.
irrelevant
items
in
a
pre-and
retro-cuing
task.
The
rationale
was
if
benefit
then
having
fewer
to
suppress
would
enhance
performance.
Instead,
found
selecting
two
out
three
slower
and
accurate
than
one,
arguing
against
idea
diminished
inhibition
underlies
efficiency.
Experiments
2a
2b
further
probed
efficiency
using
modified
dual-access
paradigm
leveraged
object
repetition
benefits.
By
including
control
condition
prevent
temporal
associations
between
repeated
targets
non-targets,
observed
benefits
for
each
item
were
additive-consistent
with
serial
limited
parallel
retrieval-rather
overadditive,
which
expected
under
fully
These
findings
clarify
key
limitations
WM,
important
implications
complex
tasks
such
language
comprehension,
decision-making,
problem
solving.