bioRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2024,
Volume and Issue:
unknown
Published: Feb. 17, 2024
Abstract
Decision-making
in
mammals
is
fundamentally
based
on
the
integration
of
several
senses.
When
different
sensory
inputs
guide
conflicting
outcomes,
animals
make
decisions
following
prepotent
modality.
However,
neural
basis
state-dependent
flexibility
dominance
and
multisensory
remains
obscure.
Here,
we
found
that
locomotion
switched
auditory-dominant
to
visual-dominant
ones
mice
with
audiovisual
conflicts.
made
decisions,
visual
representation
posterior
parietal
cortex
(PPC)
was
weakened.
Prolonged
restored
PPC
induced
visually
dominant
by
inhibiting
auditory
cortical
neurons
projecting
(AC
).
In
contrast,
it
had
no
effect
AC
projection
striatum,
which
mediated
decisions.
Locomotion
activated
AC-projecting
secondary
motor
neurons,
preferentially
inhibited
neurons.
Our
findings
demonstrate
flexible
gating
afferents
association
depending
locomotor
conditions.
Cell Reports,
Journal Year:
2023,
Volume and Issue:
42(3), P. 112246 - 112246
Published: March 1, 2023
The
ability
to
abstract
information
guide
decisions
during
navigation
across
changing
environments
is
essential
for
adaptation
and
requires
the
integrity
of
hippocampal-prefrontal
circuitry.
hippocampus
encodes
navigational
in
a
cognitive
map,
but
it
remains
unclear
how
maps
are
transformed
circuits
support
abstraction
generalization.
Here,
we
simultaneously
record
ensembles
as
rats
generalize
rules
distinct
environments.
We
find
that,
whereas
hippocampal
representational
maintain
specificity
separate
environments,
prefrontal
Furthermore,
while
both
structured
within
neural
manifold
population
activity,
they
have
geometries.
Prefrontal
geometry
enables
rule-informative
variables,
format
that
generalizes
novel
conditions
existing
variable
classes.
Hippocampal
lacks
such
abstraction.
Together,
these
findings
elucidate
into
geometric
representations
generalization
maintaining
memory
specificity.
Nature Communications,
Journal Year:
2023,
Volume and Issue:
14(1)
Published: April 14, 2023
Abstract
Decision-making
requires
flexibility
to
rapidly
switch
one’s
actions
in
response
sensory
stimuli
depending
on
information
stored
memory.
We
identified
cortical
areas
and
neural
activity
patterns
underlying
this
during
virtual
navigation,
where
mice
switched
navigation
toward
or
away
from
a
visual
cue
its
match
remembered
cue.
Optogenetics
screening
V1,
posterior
parietal
cortex
(PPC),
retrosplenial
(RSC)
as
necessary
for
accurate
decisions.
Calcium
imaging
revealed
neurons
that
can
mediate
rapid
switches
by
encoding
mixture
of
current
These
mixed
selectivity
emerged
through
task
learning
predicted
the
mouse’s
choices
forming
efficient
population
codes
before
correct,
but
not
incorrect,
choices.
They
were
distributed
across
cortex,
even
densest
RSC
sparsest
PPC.
propose
decisions
arises
mix
memory
within
visual-parietal-retrosplenial
network.
Neuron,
Journal Year:
2024,
Volume and Issue:
112(11), P. 1876 - 1890.e4
Published: March 5, 2024
In
complex
environments,
animals
can
adopt
diverse
strategies
to
find
rewards.
How
distinct
differentially
engage
brain
circuits
is
not
well
understood.
Here,
we
investigate
this
question,
focusing
on
the
cortical
Vip-Sst
disinhibitory
circuit
between
vasoactive
intestinal
peptide-postive
(Vip)
interneurons
and
somatostatin-positive
(Sst)
interneurons.
We
characterize
behavioral
used
by
mice
during
a
visual
change
detection
task.
Using
dynamic
logistic
regression
model,
that
individual
use
mixtures
of
comparison
strategy
statistical
timing
strategy.
Separately,
also
have
periods
task
engagement
disengagement.
Two-photon
calcium
imaging
shows
large
strategy-dependent
differences
in
neural
activity
excitatory,
Sst
inhibitory,
Vip
inhibitory
cells
response
both
image
changes
omissions.
contrast,
has
limited
effects
population
activity.
diversity
correlates
be
understood
parsimoniously
as
increased
activation
strategy,
which
facilitates
task-appropriate
responses.
PLoS Computational Biology,
Journal Year:
2023,
Volume and Issue:
19(9), P. e1011430 - e1011430
Published: Sept. 14, 2023
In
reversal
learning
tasks,
the
behavior
of
humans
and
animals
is
often
assumed
to
be
uniform
within
single
experimental
sessions
facilitate
data
analysis
model
fitting.
However,
agents
can
display
substantial
variability
in
sessions,
as
they
execute
different
blocks
trials
with
transition
dynamics.
Here,
we
observed
that
a
deterministic
task,
mice
noisy
sub-optimal
choice
transitions
even
at
expert
stages
learning.
We
investigated
two
sources
sub-optimality
behavior.
First,
found
exhibit
high
lapse
rate
during
task
execution,
reverted
unrewarded
directions
after
transitions.
Second,
unexpectedly
majority
did
not
strategy,
but
rather
mixed
between
several
behavioral
modes
quantified
use
such
mixtures
state-space
model,
block
Hidden
Markov
Model
(block
HMM),
dissociate
dynamic
individual
trials.
Additionally,
blockHMM
rodent
accounted
for
by
types
algorithms,
model-free
or
inference-based
learning,
might
used
solve
task.
Combining
these
approaches,
mixture
both
exploratory,
strategies
deterministic,
explaining
their
overall
sequences.
Together,
our
combined
computational
approach
highlights
intrinsic
noise
provides
richer
description
than
conventional
techniques,
while
uncovering
hidden
states
underlie
block-by-block
Nature Communications,
Journal Year:
2024,
Volume and Issue:
15(1)
Published: Aug. 1, 2024
Behavioral
flexibility
relies
on
the
brain's
ability
to
switch
rapidly
between
multiple
tasks,
even
when
task
rule
is
not
explicitly
cued
but
must
be
inferred
through
trial
and
error.
The
underlying
neural
circuit
mechanism
remains
poorly
understood.
We
investigated
recurrent
networks
(RNNs)
trained
perform
an
analog
of
classic
Wisconsin
Card
Sorting
Test.
consist
two
modules
responsible
for
representation
sensorimotor
mapping,
respectively,
where
each
module
comprised
a
with
excitatory
neurons
three
major
types
inhibitory
neurons.
found
that
by
self-sustained
persistent
activity
across
trials,
error
monitoring
gated
mapping
emerged
from
training.
Systematic
dissection
RNNs
revealed
detailed
consistent
different
hyperparameters.
networks'
dynamical
trajectories
rules
resided
in
separate
subspaces
population
activity;
collapsed
performance
was
reduced
chance
level
dendrite-targeting
somatostatin-expressing
interneurons
were
silenced,
illustrating
how
phenomenological
description
representational
explained
specific
mechanism.
flexible
switching
unclear.
Here
authors
analyzed
modular
network
models
cell
reveal
uncued
switching.
