Multivoxel Object Representations in Adult Human Visual
Cortex Are Flexible: An Associative Learning Study

Mehdi Senoussi1,2, Isabelle Berry3,4, Rufin VanRullen1,2, and Leila Reddy1,2

Astratto

■ Learning associations between co-occurring events enables
us to extract structure from our environment. Medial-temporal
lobe structures are critical for associative learning. Tuttavia, IL
role of the ventral visual pathway ( VVP) in associative learning is
not clear. Do multivoxel object representations in the VVP re-
flect newly formed associations? We show that VVP multivoxel
representations become more similar to each other after hu-
man participants learn arbitrary new associations between pairs
of unrelated objects (faces, houses, cars, chairs). Participants
were scanned before and after 15 days of associative learning.
To evaluate how object representations changed, a classifier
was trained on discriminating two nonassociated categories
(per esempio., faces/houses) and tested on discriminating their paired
associates (per esempio., cars/chairs). Because the associations were ar-
bitrary and counterbalanced across participants, there was ini-

tially no particular reason for this cross-classification decision
to tend toward either alternative. Nonetheless, after learning,
cross-classification performance increased in the VVP (but not
hippocampus), on average by 3.3%, with some voxels showing
increases of up to 10%. Per esempio, a chair multivoxel repre-
sentation that initially resembled neither face nor house repre-
sentations was, after learning, classified as more similar to that
of faces for participants who associated chairs with faces and
to that of houses for participants who associated chairs with
houses. Additionally, learning produced long-lasting perceptual
consequences. In a behavioral priming experiment performed
several months later, the change in cross-classification perfor-
mance was correlated with the degree of priming. Così, VVP
multivoxel representations are not static but become more sim-
ilar to each other after associative learning. ■

INTRODUCTION

We can rapidly and accurately detect and categorize ob-
jects even when they are flashed for just a fraction of
a second. This astonishing ability relies on the ventral
visual pathway ( VVP), a neural system that extends from
the occipital cortex to lateral and ventral regions of the
temporal lobe (Grill-Spector, 2003). The VVP is not orga-
nized in a homogenous fashion (Grill-Spector & Malach,
2004). Invece, this expanse of cortex is dotted with sev-
eral smaller regions that respond preferentially to specific
classes of stimuli (per esempio., faces, places, objects, or bodies;
Downing, Jiang, Shuman, & Kanwisher, 2001; Epstein &
Kanwisher, 1998; Kanwisher, McDermott, & Chun, 1997).
This underlying architecture is remarkably consistent
across normal, healthy participants (Haxby et al., 2011).
Object category representations in the VVP can be de-
scribed at two different levels: in the activity of large-scale
multivoxel patterns (MVPs) or at the level of the object
selectivity of individual neurons (Reddy & Kanwisher,
2006). Although it is difficult to measure the selectivity
of single neurons in the human brain, it is now well estab-
lished that object category information is also reflected in

1Université de Toulouse, 2CNRS, CerCo, Toulouse, France,
3Inserm Imagerie cérébrale et handicaps neurologiques UMR
825, Toulouse, France, 4Centre Hospitalier Universitaire de
Toulouse Pôle Neurosciences CHU Purpan

© 2016 Istituto di Tecnologia del Massachussetts

the large-scale MVPs of activity that can be recorded with
fMRI. Infatti, decoding studies have shown that category
information is explicit in these response patterns (Op de
Beeck, Brants, Baeck, & Wagemans, 2010; Reddy &
Kanwisher, 2007; Spiridon & Kanwisher, 2002; Haxby
et al., 2001). Here we ask if MVPs for well-learned catego-
ries still maintain flexibility related to visual experience in
the adult brain.

Specifically, in this study, we directly test if large-scale
representations for highly familiar categories in the VVP
become more similar to each other when pairs of catego-
ries are behaviorally associated through extensive train-
ing. At the neuronal level, anterior ventral temporal
cortex and medial-temporal lobe (MTL) structures have
been implicated in associative learning in both monkeys
( Wirth et al., 2003; Messinger, Squire, Zola, & Albright,
2001; Miyashita & Chang, 1988) and humans (Ison, Quian
Quiroga, & Fried, 2015; Reddy et al., 2015). Tuttavia,
here we show that preexisting multivoxel representa-
tions for familiar objects (faces, houses, chairs, cars) In
ventral visual cortex shift in a concerted way in a high-
dimensional multivoxel space once two categories be-
come perceptually related.

We scanned 20 observers before and after they learned
arbitrary associations between different object categories
(faces, houses, cars, chairs) and investigated changes in
the large-scale category representations with MVP analysis

Journal of Cognitive Neuroscience 28:6, pag. 852–868
doi:10.1162/jocn_a_00933

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

Figura 1. Experimental protocol and hypothesis. (UN) Each participant followed a three-step procedure. In the first step, participants performed a
prelearning scan in which they viewed blocks of faces, houses, chairs, cars, and scrambled images. Prossimo, In 15 daily sessions, participants performed a
learning task in which they learned arbitrary associations between members of the different categories. In this example, faces are paired with cars and
houses with chairs. Category pairings were counterbalanced across subjects. Each learning session consisted of 12 blocks of 40 trials. On each trial,
participants were presented with a main stimulus (per esempio., a face) and two choice stimuli from the associated category (per esempio., two cars) and had to decide which
of the choice stimuli was paired with the main stimulus (by pressing the left or right arrow keys on the keyboard). After the learning sessions, participants
performed a postlearning scan that was identical to the prelearning scan except that the block order was randomized. (B) To evaluate the similarity
between category representations before and after learning we used a cross-classification procedure with the searchlight method. An SVM classifier was
trained to distinguish between two categories and tested on their associated categories. We hypothesized that after learning, we would see an increase in
cross-classification performance suggesting that the multivoxel representations of the paired categories had become more similar to each other.

metodi. In particular, we trained a support vector
machine (SVM) classifier to discriminate between two
nonassociated object categories (per esempio., houses vs. faces)
and then tested it on discriminating between their paired
associates (per esempio., cars vs. chairs). We hypothesized that
after learning we would see an increase in this cross-
classification performance. Because a classification deci-
sion reflects the distance and the relative position of test
patterns in a multidimensional space, an increase in
cross-classification performance after learning would imply
that the representations of the paired categories had
moved in a high-dimensional multivoxel space or, equiva-
lently, had become more similar to each other.

Using this cross-classification approach, we found an
increase in decoding performance after learning, suggest-
ing that large-scale fMRI response patterns in the VVP for
associated object categories become more similar to each
other. In other words, in an example participant who as-
sociated faces with chairs and houses with cars, face
MVPs became more similar to chair MVPs and house

MVPs became more similar to car MVPs after learning.
This shift in category representations had perceptual con-
sequences, as measured by a behavioral priming task per-
formed several months after the associations had been
learned. Specifically, we found that a given category facil-
itated the processing of its paired associate relative to the
processing of a nonassociated category. Inoltre, Questo
priming effect was correlated across participants with the
overall amount by which the category representations
shifted as a result of learning.

METHODS

Participants and Stimuli

Twenty-one participants were recruited for this study
(10 women, mean age = 24 years, range = 19–35 years).
One participant was excluded from the study because
of excessive motion in the scanner. All participants had
normal or corrected-to-normal vision and reported no

Senoussi et al.

853

history of neurological problems. All participants pro-
vided written informed consent and received monetary
compensation for their participation. The local ethics
committee for human experimentation approved all
procedures.

Ten stimuli from each of four categories (faces, houses,
chairs, cars) were gathered from different sources on the
Internet. These images were then transformed to gray-
scale and pasted on a 500 × 500 pixels gray canvas. A
avoid low-level category confounds, we normalized cate-
gories in luminance, contrasto, and size. We then gener-
ated a scrambled version of each image for the functional
ROI localizers.

Experimental Protocol

The experimental protocol consisted of three phases: UN
prelearning fMRI scan, an associative learning task out-
side the scanner over 15 days, and a postlearning fMRI
scan.

During the fMRI scans, stimuli were presented with the
VisionEgg toolbox (Straw, 2008). Each fMRI run consisted
of four blocks each of the four categories (faces, houses,
cars, and chairs) and scrambled images and five blocks of
fixation. Each block was 16 sec long. The fixation blocks
occurred after every five visual stimulation blocks. In
each visual stimulation block, 16 stimuli were presented,
each for 800 msec followed by an ISI of 200 msec. Partic-
ipants were instructed to press a button when the same
image was presented on two successive trials (1-back
task). Each fMRI session consisted of eight runs that
lasted approximately 6 min and 45 sec each. The pre-
and postlearning fMRI sessions were identical, except
for the block and stimulus order, which were randomized
in each run.

In between the fMRI sessions, participants underwent
15 daily learning sessions during which they learned as-
sociations between exemplars of the object categories
(per esempio., each face was associated with a given car, and each

house with a given chair). Each 20-min session consisted
Di 12 blocks of 40 trials. Each trial lasted up to 3 sec with
an intertrial interval of 0.750 sec. On each trial, partici-
pants were presented with a main stimulus (per esempio., a chair)
and two choice stimuli (per esempio., two houses) and had to de-
cide (by pressing one of two keys on the keyboard)
which of the choice stimuli was the correct associate of
the main stimulus (Figure 1A). Exemplars of each cate-
gory served as the main stimulus or choice stimuli on dif-
ferent blocks. Ten exemplars per category were used.
Learning was achieved by trial and error, and negative
auditory feedback was provided on incorrect trials. IL
category pairings were counterbalanced across partici-
pants: Half of the participants associated faces with cars
and houses with chairs, and the other half associated
faces with chairs and houses with cars.

Priming Experiment

The priming experiment was performed on average
14.1 months after the postlearning fMRI scan on 14 Di
the original 20 participants. Before participants per-
formed the priming experiment, they underwent three
training sessions on the main associative learning para-
digm. They then performed two sessions of the priming
experiment on two days.

To avoid low-level priming effects, we equalized all
stimuli in the Fourier amplitude spectrum. On each trial
of the priming experiment, participants were presented
with a prime stimulus for 100 msec followed by a target
stimulus for 2 sec and instructed to report the category of
the target stimulus (Figura 2). The intertrial interval was
1000 msec, with a jitter of 500, 750, O 1000 msec. Dopo
each trial, the fixation cross turned to a dash for 1 sec and
turned back to a cross to signal the beginning of the next
trial. The prime and target stimuli were exemplars of the
four object categories (faces, houses, chairs, cars). Within
a block of trials, only two categories were targets (per esempio., cars
and chairs in blocks when participants were asked to

Figura 2. Behavioral priming task experiment design: On each trial of the priming experiment, participants were presented with a prime stimulus
followed by a target stimulus and instructed to report the category of the target stimulus. The prime and target stimuli were exemplars of the four object
categorie (faces, houses, chairs, cars). There were four types of trials: When the primes and targets were different exemplars from the same category
(“same” trials), when the prime and target were from opposite categories with respect to the category discrimination task (“opposite” trials), E
when the prime and target were from associated/nonassociated categories.

854

Journal of Cognitive Neuroscience

Volume 28, Numero 6

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

discriminate cars from chairs), whereas all four categories
could serve as primes. There were four types of trials:
when the prime and target category matched (“same” tri-
COME), when the prime and target were from opposite cate-
gories with respect to the category discrimination task
(per esempio., the prime was a car and the target was a chair in a
block when participants were instructed to discriminate
cars from chairs, “opposite” trials), and when the prime
and target were from associated/nonassociated categories
(per esempio., faces/houses, “associated”/“nonassociated” trials).
For “associated” trials, the primes and targets were the
pairs learned during the associative learning paradigm,
Per esempio, a participant who had learned to associate
face1 with chair5 was presented with face1 as a prime
when chair5 was the target on “associated” trials, in a block
where participants were instructed to discriminate cars
from chairs. Participants were instructed to respond as fast
as possible on each trial. Each participant performed eight
blocks of 250 trials. Trials were randomized within each
block. Participants performed the priming experiment
Sopra 2 days. On the first day, the targets were cars and
chairs, each with their own response button (left and
right, rispettivamente). On the second day, the targets were
faces and houses, each with their own response button
(up and down, rispettivamente). We chose this design to
avoid confusing participants by switching instructions
within a single experiment session.

fMRI Data Acquisition and Analysis

fMRI data were collected on a 3T Philips (Amsterdam, IL
Netherlands) ACHIEVA scanner (gradient-echo pulse se-
quence, repetition time = 2 sec, echo time [TE] = 35 msec,
30 slices with a 32-channel head coil, slice thickness =
2 mm, in-plane voxel dimensions 1.88 × 1.88 mm). IL
slices were positioned to cover the entire temporal and
occipital lobes. High-resolution anatomical images were
also acquired per participant (1 × 1 × 1 mm voxels, repe-
tition time = 8.13 msec, TE = 3.74 msec, 170 sagittal
slices). Data analysis was performed with FreeSurfer and
the FreeSurfer Functional Analysis Stream (FS-FAST; surfer.
nmr.mgh.harvard.edu), custom Matlab scripts, and the
PyMVPA toolbox (www.pymvpa.org/; Hanke et al., 2009).
Similar results were also obtained with the SearchMight
Toolbox (www.princeton.edu/∼fpereira/searchmight/).

Preprocessing followed the FS-FAST processing
stream. All images were motion-corrected (using AFNI
with standard parameters), slice time-corrected, intensity-
normalized, and smoothed with a 3-mm FWHM Gaussian
kernel. We then estimated the beta weights using a general
linear model (GLM) for the five stimulus conditions (faces,
houses, cars, chairs, and scrambled) in each participant.
The betas were computed on whole-run data. There were
eight runs in each scan session and four blocks of 16 sec of
each condition in each run. We obtained eight beta images
per condition (cioè., one from each run) from each scann-
ing session from the FS-FAST processing stream. IL

GLM fit the hemodynamic response with a gamma function
(delta = 2.25, tau = 1.25) and modeled the drift with an
order 1 polynomial. For all other parameters of the GLM,
we used the default settings from FS-FAST. Finalmente, IL
beta-weight volumes were normalized on the MNI305
brain, and we used these volumes as inputs for the search-
light analysis. Similar results were obtained when the
searchlight analysis was performed in the native space of
each participant.

ROIs
ROIs were defined manually in each participant’s native
space using an independent analysis. Fusiform face area
(FFA) was defined as the set of contiguous voxels in the
fusiform gyrus that exhibited greater activation for faces
than houses ( P < 10−5, uncorrected). Parahippocampal place area (PPA) was defined as the set of contiguous voxels in the parahippocampal gyrus that exhibited greater acti- vation for houses than faces ( p < 10−5, uncorrected). lateral occipital complex (LOC) was defined as the set of voxels in the inferior occipital and temporal cortices that exhibited greater activation for cars and chairs than scrambled images ( p < 10−5, uncorrected). The anterior and posterior subdivisions of LOC (lateral occipital [LO] and posterior fusiform [pF]) were also identified for each participant. The hippocampus, V1, and V2 were defined using anatomical landmarks for each participant in FreeSurfer. The average ROIs displayed in Figure 7 were computed by selecting voxels that were common to at least 60% of the ROIs defined in individual partici- pants. Note that the ROI analyses were performed in each participant’s individual ROIs, and the average ROI is used for display purposes only. Multivariate Analysis The searchlight analysis was performed with the CrossValidation, HalfPartitioner, LinearCSVMC, and sphere_ searchlight functions of the PyMVPA toolbox using default settings. A linear SVM with default settings from the PyMVPA toolbox was used to perform a cross-classification analysis within each searchlight. We used searchlights of different radii (1–14 voxels) that we moved along the MNI305 vol- umes. For each participant, within each searchlight, an SVM classifier was trained on the fMRI patterns for two nonassociated categories for that participant (e.g., faces vs. houses) and tested on the corresponding associated categories (e.g., cars vs. chairs). Additionally, the symmet- ric classification was also performed (i.e., in the example here, a car–chair classifier was tested on a face–house dis- crimination). The average of the two classification scores was reported as the cross-classification performance for the voxel at the center of the searchlight. The input to the classifiers were eight MVPs for each condition. For example, when training a classifier on a face versus house discrimina- tion and testing it on a car versus chair discrimination, the Senoussi et al. 855 l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / j / o c n a r t i c e - p d l f / / / / 2 8 6 8 5 2 1 7 8 4 9 5 7 / j o c n _ a _ 0 0 9 3 3 p d . f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 classifier was trained on eight MVPs of face betas and eight MVPs of house betas and tested on eight MVPs of car betas and eight MVPs of chair betas. The same stimuli and data sets were used for the ex- perimental sessions and for defining the functional ROIs. However, our analysis is free of the double-dipping prob- lem (Kriegeskorte, Simmons, Bellgowan, & Baker, 2009) because orthogonal contrasts were used in defining the ROIs versus in the cross-classification analysis. For in- stance, when defining the FFA, we used a faces–houses contrast. On the other hand, the cross-classification anal- ysis tested a face/house classifier on a car/chair discrimi- nation. Defining our FFA with a face–house contrast guarantees that a face/house classifier in these voxels would perform superbly on a face/house discrimination of the same data (i.e., a circular analysis). However, there is no reason for performance of the face/house classifier on a car/chair discrimination task to benefit from this method of voxel selection. Correct or incorrect classification depended on the asso- ciation learned by the particular participant. For example, for participants who learned face–car and house–chair as- sociations, cross-classification would be deemed as correct if the face–house classifier classified the MVP elicited by cars as faces and the MVP elicited by chairs as houses. This procedure produces cross-classification performances ranging from 0 to 1. A performance of 1 means that the classifier always considered patterns of associated catego- ries as being more similar, a performance of 0 means that it always considered patterns of nonassociated categories as being more similar, and a score of 0.5 means that the classifier did not have any bias between the categories. This procedure could be done in two ways, because there were two pairs of associations: training the classifier on faces and houses and testing it on cars and chairs, or training it on cars and chairs and testing it on faces and houses. The re- sults of these two analyses were equivalent so the final cross-classification performance values were averaged across the two analyses. Note that the cross-classification approach might be a more sensitive test of learning-induced flexibility than a direct classification test on the associated category pairs because a priori, a chair pattern should fall roughly half- way between a face and a house pattern (i.e., 50% classi- fication performance), so a small shift of the chair pattern toward the face pattern could result in a sizeable change in cross-classification performance. On the other hand, if face and chair patterns become more similar in a multi- dimensional space, they might still be far enough apart that a direct face/chair classifier would never confuse a chair with a face and thus learning would not seem to modify classification accuracy. The searchlight analysis was performed across the en- tire scanned functional volume as well as in the specific l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / j / o c n a r t i c e - p d l f / / / / 2 8 6 8 5 2 1 7 8 4 9 5 7 / j o c n _ a _ 0 0 9 3 3 p d . f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Figure 3. Behavioral results during learning. Each participant performed 15 learning sessions outside the scanner. The RTs and accuracies in each session are shown here (individual sessions indicated by the blue and white areas). RTs (top plot) decreased steadily (one-way, random-effects ANOVA on log(RT): F(14, 266) = 128.61, p < 10−6), and stabilized after the tenth session (post hoc Tukey’s HSD ( p < .05)). For statistical tests only (but not for display purposes), the RTs were log-transformed to satisfy the constraints of normality. Accuracy (bottom plot) was computed for each session as the proportion of trials where the participant responded correctly. Accuracy on the learning task was at or above 90% by the end of the first learning session for most participants (19 of 20) and then stabilized by the second session (one-way, random-effects ANOVA: F(14, 266) = 9.38, p < 10−6, post hoc Tukey’s HSD ( p < .05)). The pink lines correspond to the average across participants of all trials of each block in each session, and the shaded area is the SEM. 856 Journal of Cognitive Neuroscience Volume 28, Number 6 ROIs defined for each participant. For the ROI-specific analyses, we used a fixed-size searchlight of radius 3 vox- els (i.e., a searchlight consisting of 123 voxels). Statistical Analysis The statistical significance of the difference between the pre- and postlearning distributions was evaluated using two-tailed one-sample t tests across participants. To as- sess the statistical significance of the voxels that showed the largest cross-classification shifts in Figure 4 (and in the corresponding surface maps in Figure 7), we used a nonparametric test in which we shuffled the labels of the pre- and postlearning sessions for each voxel and for each participant independently to simulate the null hy- pothesis that there was no difference between these ses- sions for each voxel. The surrogate distributions were computed 2000 times per participant. The p value of each voxel was assigned by comparing this voxel’s cross-classification shift to the corresponding surrogate values (i.e., 2000 iterations × 78,842 voxels). Correlation Analysis In the pF, FFA, PPA, and LO, a searchlight of radius 3 vox- els, centered on each voxel, was trained and tested on discriminating the four categories (faces, houses, chairs, cars) from each other prior to learning. This four-way classification analysis was performed individually for each participant in the MNI305 space and then averaged across participants to obtain an average four-way classifi- cation performance value for each voxel. This average four-way classification performance value was then corre- lated with the change in cross-classification performance of each voxel (also averaged across participants to obtain one performance value per voxel). The parameters of the four-way classifier were identical to the cross-classification classifier (see above). The four-way classifier was trained on blocks of data from all runs but one and tested on the remaining run (leave-one-run-out cross-validation). This correlation analysis was performed on the average ROIs computed by selecting voxels that were common to at least 60% of the ROIs defined in individual participants. To make the correlations comparable across ROIs, we equalized the number of voxels in each ROI before com- puting the correlation value. Specifically, we resampled each ROI 100,000 times, each time randomly choosing 162 voxels (that corresponded to the size of the smallest ROI, pF) and computing the Pearson’s r value in each re- sample. The reported r2 values correspond to the square of the average r values of these resamples. RESULTS Twenty observers were scanned before and after they learned arbitrary associations between pairs of different object categories (Figure 1A). During these pre- and postlearning fMRI scans, the participants viewed 10 ex- emplars each of faces, houses, chairs, cars, and scrambled images in different blocks. Participants performed a 1-back task, in which they responded if the same image had been presented on two successive trials. Note that, in the scan- ner, the image presentation order and the 1-back behav- ioral task were independent of the associations learned by the participants outside the scanner. These scans simply allowed us to obtain pre- and postlearning MVPs for the four object categories. In between these two scan sessions, participants learned arbitrary associations between different exem- plars of the four object categories (e.g., each face was as- sociated with a car/each house with a chair; Figure 1A). Most participants achieved greater than 90% accuracy by the end of the first session and continued to improve until behavioral measures of learning stabilized by the tenth session. Participants continued to train even after performance had stabilized (Figure 3). As mentioned above, we trained an SVM classifier to discriminate between two nonassociated categories (e.g., faces and houses) and tested it on discriminating their paired associates (e.g., cars vs. chairs). We hypoth- esized that after learning we would see an increase in this cross-classification performance (Figure 1B), suggesting that the multivoxel representations of the paired catego- ries had become more similar to each other. Because we had no strong a priori expectation about where these learning-related changes might occur, we used the search- light method to explore different areas of the VVP (Kriegeskorte, Goebel, & Bandettini, 2006). To perform the cross-classification procedure, we rea- ligned each participant’s functional volume to the MNI305 brain to make comparisons across participants. We moved a spherical searchlight along each participant’s realigned functional volume and, at each voxel, calculated the cross-classification performance from the MVPs falling within the searchlight centered on that voxel. More specif- ically, for an example participant who had learned to asso- ciate faces with cars and houses with chairs, we tested the performance of a face–house classifier on car–chair dis- crimination and a car–chair classifier on face–house dis- crimination within the searchlight. Note that there is no “correct” answer for either of these classifiers, as the asso- ciations were arbitrarily determined—we simply assumed that, faced with a meaningless choice, the classifier would tend to choose the label of the associated category. The average of these two classification scores was the cross- classification score attributed to the voxel at the search- light center. We performed this analysis separately on the MVPs from the pre- and postlearning scans and evalu- ated how cross-classification performance changed after learning. The pre- and postlearning distributions of cross- classification performance across all the voxels in the scanned volume were averaged across the 20 participants and are shown in Figure 4A. To obtain optimal classification Senoussi et al. 857 l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / j / o c n a r t i c e - p d l f / / / / 2 8 6 8 5 2 1 7 8 4 9 5 7 / j o c n _ a _ 0 0 9 3 3 p d . f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / j / o c n a r t i c e - p d l f / / / / 2 8 6 8 5 2 1 7 8 4 9 5 7 / j o c n _ a _ 0 0 9 3 3 p d . f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Figure 4. Cross-classification performance before and after learning. (A) Histograms showing the distribution of cross-classification performance across all voxels in the scanned volume, obtained with a searchlight of radius 12 voxels, averaged across 20 participants. Prelearning performance values are in blue, and postlearning performance values are in pink. The overlap between the pre- and postlearning distributions is shown in purple. Across all voxels, there was a significant increase in performance after learning (3.3 ± 0.95%; (t(19) = 3.39, p < .005). (B) Distribution of the voxelwise difference between the pre- and postlearning performance values. As expected from A, the average voxelwise difference was 3.3%. However, although the shift was absent or only moderate for some voxels, a number of voxels shifted by more than 10% on average. (C) The effect size of the difference between pre- and postlearning cross-classification distributions obtained with searchlights of different radii. (D–G) Same as in A for different ROIs, obtained with a searchlight of radius 3 voxels. The significance and shift of the difference between the pre- and postlearning distributions are indicated for each panel. performance, the size of the searchlight must be com- mensurate with the size of the region where the effects occur (Kriegeskorte et al., 2006). Accordingly, we tested the effect of varying the searchlight radius on cross- classification performance. In the whole-brain analysis we obtained optimal cross-classification performance with a searchlight of radius 12 voxels (Figure 4A), but similarly significant effects were also obtained with searchlights of other radii from 1 to 14 voxels (Figure 4C). The pre- and postlearning average cross-classification performance values were 48.3% and 51.7%, respectively, and not signif- icantly different from chance levels (50%, t(19) = 1.55, p = .13 for the prelearning distribution and t(19) = 1.82, p = .08 for the postlearning distribution). However, between the two learning sessions, we observed a significant in- crease of 3.3 ± 0.95% (mean ± SEM ) in the average cross-classification performance over all voxels in the scanned volume. We statistically evaluated this difference in mean cross-classification performance between the two scan sessions using a two-tailed, paired t test of average pre- and postlearning performances (with each participant contributing one global cross-classification performance value to the statistical test, thus avoiding multiple com- parisons across voxels or brain regions and warranting the assumption of independence between measurements (t(19) = 3.39, p < .003). Furthermore, the increase in cross-classification performance after learning was not driven by the type of association learned by participants (Figure 5): Similar increases were observed for the par- ticipants who had associated faces with cars and houses with chairs (increase of 3.4 ± 1.4%, two-tailed, paired t test; t(9) = 2.26, p = .05) as for the participants who had associated faces with chairs and houses with cars (in- crease of 3.2 ± 1.2%, t(9) = 2.42, p < .04). 858 Journal of Cognitive Neuroscience Volume 28, Number 6 As argued above, the shift in the distribution of cross- classification performance suggests that multivoxel object category patterns become more similar to each other af- ter participants learn associations between the catego- ries. How sparse are these learning-related changes? On the one hand, the representational changes could poten- tially be highly variable across voxels, with voxels in some areas showing a large increase in performance after learn- ing and others showing no change at all. Alternatively, at the other extreme, every voxel in the scanned volume could shift by the same amount. To determine how spe- cific the learning-induced changes were, we evaluated the distribution of voxelwise differences between the pre- and postlearning classification performances (Fig- ure 4B). As expected from the results in Figure 4A, the average voxelwise increase in cross-classification per- formance after learning was 3.3%. However, the shift was variable: Some voxels showed an increase in cross- classification performance of more than 10%. Figure 6 shows the scatter of pre- and postlearning cross-classification performances across all voxels, and its relationship to the histograms shown in Figure 4A and B. We next asked how the voxels that showed the largest shifts in performance were organized in cortex. In other words, did they occur together in localized regions or were they dispersed all across cortex? Some authors have suggested that expertise-related changes might occur in specific ROIs, for example, in the fusiform gyrus (Gauthier, Tarr, Anderson, Skudlarski, & Gore, 1999). In a first step, we thus evaluated how learning affected object represen- tations in functionally defined regions of ventral temporal cortex that are known to be important for processing visual categories. In particular, we identified four functionally defined regions in each participant’s native space: the FFA (Kanwisher et al., 1997), the PPA (Epstein & Kanwisher, 1998), and the pF and LO subdivisions of the LOC (Grill- Spector, Kourtzi, & Kanwisher, 2001). In addition, we ana- tomically identified the hippocampus (because of its implication in the acquisition of new associations), and the early visual cortex (V1 and V2) as a control region. In each of these areas, we performed the same analysis as in Figure 4A (Figure 4D–G). However, because we were considering smaller ROIs, we restricted this analysis to a smaller searchlight of radius 3 voxels. Note that we retained l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / j / o c n a r t i c e - p d l f / / / / 2 8 6 8 5 2 1 7 8 4 9 5 7 / j o c n _ a _ 0 0 9 3 3 p d . f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Figure 5. Cross-classification performance by association type. In our group of 20 participants, half the participants associated faces with cars and houses with chairs (Group 1; A), whereas the other half associated faces with chairs and houses with cars (Group 2; B). Both groups of participants showed similar effects of associative learning (independent samples t test, t(9) = 0.11, p = .9). A and B have the same format as in Figure 4A. C and D have the same format as in Figure 4B. Senoussi et al. 859 shifted significantly ( p < .001, uncorrected) corre- sponded to a performance shift of at least 9.25% and were clustered principally in the left and right fusiform gyri. In particular, the largest group of these voxels over- lapped with the functionally defined left and right FFA and an anterior subdivision of the LOC known as pF (Grill-Spector et al., 1999). As can be seen in Figures 4 and 7, the learning-induced changes were not uniform within each ROI. Instead, some voxels exhibited greater shifts in cross-classification performance than others. We next investigated what characterized those voxels that showed higher flexibility. We reasoned that flexibility might be inversely related to initial selectivity, that is, that the voxels that originally provided the most information about object category (i.e., the most specialized voxels) would retain their se- lectivity, whereas the least informative voxels would be most sensitive to category associations during the learn- ing phase. Thus, in each of the previously identified ROIs (pf, FFA, PPA, and LO), we compared the voxelwise in- crease in cross-classification performance after learning with the ability of that voxel to provide category-specific information before learning (i.e., the performance of a classifier trained on a set of patterns and tested on pat- terns from the same category). For each voxel, the per- formance of a four-way classifier (3-voxel radius spherical searchlight centered on that voxel), trained and tested on discriminating the four categories (faces, houses, chairs, Figure 7. Localization of voxels that showed the largest increase in cross-classification accuracy. The voxels that showed the largest increase in cross-classification accuracy across all participants after learning were in relatively localized regions of the left and right fusiform gyri ( p < .001 uncorrected; corresponding to an increase in cross-classification accuracy of 9.25% or more). The colorbar corresponds to p values (uncorrected) determined from a nonparametric test. The outline of the functionally defined FFA (averaged across participants) is shown in blue, and the average pF is shown in green. Figure 6. Scatter plot of voxelwise cross-classification performance in the prelearning versus postlearning scan sessions. The blue and pink histograms are the projections of the data on the y- and x-axes, respectively, and are similar to the data shown in the blue and pink histograms of Figure 4A (save for the fact that here the data points and corresponding histograms represent mean classification performance of each voxel across participants). The green histogram corresponds to the data in Figure 4B and is the projection of the data perpendicular to the diagonal. a higher-resolution searchlight approach rather than test- ing a whole ROI classifier, because it is conceivable that over an entire ROI the most informative voxels (i.e., those that will dominate the classifier’s decision) may not be those that show the strongest learning effect (and indeed, this possibility was confirmed in a subsequent analysis; see Figures 8 and 9). In that case, a whole ROI clas- sifier may not show any learning-induced change in cross-classification (Figure 9), although individual voxels within the corresponding ROI could have significantly altered their response pattern; the searchlight method, on the other hand, would still reveal the changes in those voxels (Figure 4). We observed a statistically significant increase in cross-classification performance in all ROIs (pF: 5.6 ± 1.2%; t(19) = 4.33, p < .0005; FFA: 4.9 ± 1.1%; t(19) = 4.24, p < .0005, PPA: 3.8 ± 1.6%; t(19) = 2.3, p < .05, LO: 2.5 ± 0.8%; t(19) = 2.99, p < .01), but not in the hippocampus (0.2 ± 0.8%; t(19) = 0.33, p >
.7) and V1/ V2 (0.4 ± 1.2%; T(19) = 0.72, p > .36).

In a complementary analysis, we asked where the vox-
els that showed the largest increase in performance were
localized. Figura 7 shows the average voxelwise perfor-
mance differences (obtained with a searchlight of radius
3 voxels) projected on the inflated brain. To assess the
statistical significance of the cross-classification shifts,
we used a nonparametric test in which we shuffled the
labels of the pre- and postlearning sessions for each voxel
independently to simulate the null hypothesis that there
was no difference between these sessions. Voxels that

860

Journal of Cognitive Neuroscience

Volume 28, Numero 6

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

Figura 8. Voxelwise correlation of changes in cross-classification performance (Session 2–Session 1) with the initial performance of a four-way
classifier. A four-way classifier (chance at 25%) was trained and tested on prelearning data to discriminate the four object categories from each other
(cioè., trained on a set of patterns and tested on patterns of the same categories). Its performance was correlated with the learning-induced changes in
cross-classification performance for each voxel in the (UN) pF, (B) FFA, (C) PPA, E (D) LO. Increase in cross-classification performance as a result of
learning was significantly negatively correlated with the category discrimination performance in area pF ( P < .0005). The gray points correspond to the individual voxels in each ROI on which correlations were computed. For visibility only, the voxels were split into quartiles according to four-way classification performance. The mean performance for each quartile is shown by the black points (error bars correspond to SD across voxels). cars) during the first fMRI recording session (prelearn- ing), was correlated with the learning-induced change in cross-classification performance. Consistent with our hypothesis, increase in cross-classification performance was significantly negatively correlated with the initial per- formance of the four-way classifier in area pF (Figure 8). This finding indicates that, in this ROI, the voxels that ex- hibited the most flexibility during the learning procedure were the ones with the lowest category-specific informa- tion prelearning (albeit four-way classification perfor- mance in these voxels was much higher than the 25% chance level; Figure 8A). Note also that, although in pF Figure 9. Prelearning and postlearning SVM classification performance in the FFA, PPA, LO, and pF, performed at the level of the entire ROI, that is, without a searchlight method. “Standard classification” refers to the average performance of a face–house (FH) classifier tested on FH discrimination and a car–chair (CC) classifier tested on CC discrimination (using a leave-one-run-out approach). “Cross-classification” refers to the average performance of the FH classifier on CC discrimination and the CC classifier on FH discrimination. Senoussi et al. 861 l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / j / o c n a r t i c e - p d l f / / / / 2 8 6 8 5 2 1 7 8 4 9 5 7 / j o c n _ a _ 0 0 9 3 3 p d . f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Figure 10. Behavioral performance on a priming task. Participants performed a priming task and were instructed to prioritize response times over accuracy. (A) RTs were significantly shorter (21.4 ± 5.4 msec, mean ± SEM, two-tailed, paired t test; t(13) = 4.47, p < .001) for the associated categories versus the nonassociated categories. The magnitude of the priming effect on RTs was approximately 32% of the maximal priming that could be observed between “same” and “opposite” trials (66.3 ± 5.0 msec). For each participant, the mean RT across all conditions was subtracted from each condition to obtain the centered RT displayed here. For statistical tests only (but not for display), the RT data were log-transformed to satisfy the constraints of normality. (B) Although RTs were our main dependent variable, a compatible difference was also observed for accuracies on “associated” versus “nonassociated” trials (2.5 ± 0.9%, two-tailed, paired t test; t(13) = 2.53, p < .03). (C) The priming effect on RTs (expressed relative to the maximal priming) was statistically correlated over the group of participants with the difference in cross-classification performance before and after learning (r2 = .41; p < .05). the voxels with the least category selectivity showed the largest learning effects, at the level of the entire ROI pF itself was highly category selective (Figure 9). One may question the validity of using a local selectiv- ity measure (the searchlight method) to draw global con- clusions about the entire ensemble of recorded voxels across occipital and temporal cortex: global measure- ments (such as a classifier trained and tested on the en- tire set of voxels) may appear more appropriate for that purpose. In fact, however, the searchlight method al- lowed us to obtain a global measure of learning over the whole brain (by averaging across voxels) and to then subsequently hone in on more localized effects. Note that, as opposed to this approach of the searchlight method, a classification analysis performed over all the voxels in the entire scanned volume (or even in a specific ROI, as alluded to above; see Figure 9) could potentially fail to find the voxels that show the biggest changes in cross-classification. For example, a global classifier trained to discriminate faces versus houses across a large swath of cortex would identify the voxels that are the most informative (i.e., category selective) for this face/ house discrimination task and disregard the voxels that are the least category selective. However, as we observed in Figure 8, the voxels that were the most prone to learn- ing in pF (i.e., showing the most significant learning effects in a cross-classification task) were precisely the ones that were the least category selective. Thus, al- though a whole-brain classifier might assign negligible weights to these voxels and consequently fail to identify learning effects, the searchlight method would not be- cause it is constrained to learn from local patterns. Do the multivoxel representational shifts have percep- tual consequences at the behavioral level? In a priming task, performed ∼14 months after the associative learn- ing had occurred, we investigated whether perception of one category facilitated the behavioral processing of its associated category, relative to its nonassociated cate- gory (Figure 2). An examination of participants’ behav- ioral performance revealed that RTs were significantly shorter (two-tailed, paired t test; t(13) = 4.47, p < .001) on trials when the prime stimulus was a paired associate versus a nonassociate (Figure 10A). The average magnitude of this priming effect (21.4 ± 5.4 msec, mean ± SEM ) was approximately 32% of the maximal priming (66.3 ± 5.0 msec, mean ± SEM ) that could be observed between “same” and “opposite” trials. Although RTs were our main dependent variable (because partici- pants were explicitly instructed to prioritize response speed over accuracy), a compatible difference was also present for accuracies on “associated” versus “nonassoci- ated” trials (2.5 ± 0.9%, two-tailed, paired t test; t(13) = 2.53, p < .03), with a priming effect for associated cate- gories that was 22.5% of the corresponding maximal priming (Figure 10B). Could this priming effect represent a behavioral corre- late of the cortical representational shifts observed in the fMRI? In support of this idea, we found that the priming effect on RTs (expressed relative to the maximal priming) was statistically correlated over the group of participants with the difference in cross-classification performance before and after learning (r 2 = .41 p < .05; 95% confi- dence interval: .07 ≤ r 2 ≤ .75; Figure 10C). In other words, the participants who had displayed the maximal shifts in multivoxel representations were also those who showed the largest priming effects. Thus, we found that even several months after the associative learning had occurred, changes in neural representations of the associated categories were accompanied by significant and commensurate response priming at the behavioral 862 Journal of Cognitive Neuroscience Volume 28, Number 6 level (although, as with all effects based on a correlation analysis, these results cannot provide evidence for a di- rect link between the changes observed in fMRI and the behavioral priming effects). DISCUSSION In this study, we asked how associative learning changes large-scale multivoxel representations in ventral temporal cortex. After learning, we observed an average increase of 3.3% in cross-classification performance of multivoxel cat- egory representations, with some voxels showing shifts of up to 10%. Because our experiment used a block de- sign, it remains an open question whether these multi- voxel category patterns arise spontaneously in the brain under different testing regimes (Kriegeskorte, Mur, & Bandettini, 2008; Kriegeskorte, Mur, Ruff, et al., 2008). Nevertheless, our results suggest that in conditions where category-specific MVPs can be recorded, the multi- voxel representations for associated categories in object- selective cortex become more similar to each other after associations are learned. In a behavioral experiment, we verified the perceptual consequences of the shifts in mul- tivoxel representations several months after the learning had occurred. Not only did paired associates produce sig- nificant cross-category priming, but also, the participants who had displayed the maximal shifts in multivoxel rep- resentations were those who showed the largest priming effects. Note however that we cannot exclude other fac- tors that might also have contributed to the significant correlation between fMRI effects and behavior, for exam- ple, participants’ motivation levels and their ability to fol- low task instructions. Cross-classification performance after learning was sig- nificantly higher than before learning. However, when averaged over all participants and voxels (Figure 4), neither the pre- nor postlearning cross-classification performance values (48.3% and 51.7%, respectively) were significantly different from chance level (50%; t(19) = 1.55, p = .13, for the prelearning distribution and t(19) = 1.82, p = .08, for the postlearning distribution). We believe that the initially low performance value was caused by sponta- neous biases in category associations occurring in many brain areas. In the PPA, for example, on average about 62% of the chair-category patterns tended to be sponta- neously associated with house (rather than face) patterns and cars with faces (rather than houses). Of course, the counterbalanced set of participants was designed to mini- mize the effects of any such initial bias (because for one half of the participants, this bias would result in lower- than-chance prelearning cross-classification and higher than chance for the other half ). However, in our limited participant population, it is not altogether surprising that the initial bias of a few participants could have been overly represented in the grand average (e.g., because of a higher signal-to-noise ratio during scanning or be- cause the relative volume of specific ROIs was bigger in these participants), leading (in our case) to an average prelearning cross-classification below 50%. If we take this initially low value as the chance level (or baseline) for postlearning cross-classification, therefore, we observe a truly significant ( p < .003) cross-classification improve- ment due to learning. It must also be emphasized that our findings are not contingent on below-chance prelearn- ing cross-classification: Similar learning-induced improve- ments were registered for several brain regions and participants whose cross-classification accuracy started off above chance. This can be easily visualized in Figure 6: Even the voxels with the highest initial cross-classification performance demonstrated a learning-induced improve- ment (i.e., a shift to the right of the diagonal). The shifts observed in the MVPs could reflect different mechanisms by which object representations change as a result of learning. For instance, the new patterns could reflect a link (or a coactivation) between the two (un- changed) initial representations of the associated catego- ries or signal entirely new representations that combine information about the associated categories. Although it would be interesting to compare what category informa- tion is encoded in the initial versus changed representa- tions, we must note that any comparison of MVPs across the two sessions (i.e., training on patterns from one ses- sion and testing on patterns from the other) would con- found learning effects with pattern and classification differences that are simply due to the fact that the two scan sessions were obtained on different days. However, the finding that, in area pF, the voxels that showed the greatest flexibility during learning were the ones that originally provided the least (albeit still much greater than the 25% chance level) category-specific information (Figure 8) suggests that the voxels that are the most in- formative in encoding category information mostly preserve their response profiles whereas the least informative voxels are more readily modulated by associative learning. Object representations in the VVP can be described both at the level of individual neuronal selectivities as well as in large-scale multivoxel activation patterns (Reddy & Kanwisher, 2006). Indeed, in the human brain, MVPs are often used as a proxy for understanding the neuro- nal codes underlying object representation (Stansbury, Naselaris, & Gallant, 2013; Kriegeskorte, Mur, Ruff, et al., 2008; Haynes & Rees, 2005; Kamitani & Tong, 2005; Carlson, Schrater, & He, 2003; Spiridon & Kanwisher, 2002; Haxby et al., 2001). As explained earlier, the ob- served increase in cross-classification performance after learning can be described in mathematical terms as a shift of the MVPs in a high-dimensional multivoxel space. How- ever, this shift of MVPs could arise from different mecha- nisms at the neuronal level, and we can only speculate here about such neuronal properties. For instance, individ- ual neurons within each voxel could change their tuning curve as a result of learning, such that initially face-selective neurons (for example) would also now respond to the as- sociated chairs (Ison et al., 2015; Reddy et al., 2015). Such a Senoussi et al. 863 l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / j / o c n a r t i c e - p d l f / / / / 2 8 6 8 5 2 1 7 8 4 9 5 7 / j o c n _ a _ 0 0 9 3 3 p d . f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 change in tuning is equivalent at the neuronal level to the coactivation account alluded to above. That is, when partic- ipants view one stimulus (e.g., a chair), neurons that are normally selective to the associated stimulus (e.g., a face) could also be partially and automatically activated, occasion- ing a change of their tuning curve. In turn, this would im- ply that the recorded MVP in response to viewing a chair is composed of a combination of chair and face represen- tations. Alternatively, the newly learned associations could be encoded within each voxel by a new set of neu- rons that were previously nonselective for either stimulus of the associated pair. In other words, when viewing a face or a chair, in addition to the original populations of face- and chair-selective neurons (respectively), a new sub- population of neurons encoding the face–chair relation would also be activated. Although our data do not allow us to tease apart these different (and nonexclusive) mech- anisms at the neuronal level, they do provide evidence that object representations as measured by MVPs are not static. Recent studies have shown that multivoxel rep- resentations of objects in ventral temporal cortex are not fixed but can be modulated by top–down signals such as task goals (Harel, Kravitz, & Baker, 2014). Our findings add to this body of work and show that object represen- tations of highly familiar categories can flexibly move in a high-dimensional multivoxel space as a result of associa- tive learning. During tasks of explicit memory recall, when par- ticipants learn to pair two stimuli together (e.g., a word and a scene), the presentation of a cue stimulus (e.g., the word) can reactivate the fMRI representation of the associated stimulus (Gordon, Rissman, Kiani, & Wagner, 2014; Kuhl & Chun, 2014; Kuhl, Rissman, Chun, & Wagner, 2011; Johnson, McDuff, Rugg, & Norman, 2009; Polyn, Natu, Cohen, & Norman, 2005; Nyberg, Habib, McIntosh, & Tulving, 2000; Wheeler, Petersen, & Buckner, 2000). This reactivation of associated stimuli during ex- plicit recall appears to resemble the results reported here and is compatible with the coactivation account discussed above. Note, however, that this study differs in one crucial aspect from past studies of explicit recall. In the studies cited above, the reactivation of the asso- ciated MVP occurred as the participants were explicitly instructed to perform a recall task (and thus retrieve the corresponding stimulus in memory). In contrast, in our study, participants were not instructed to perform a re- call task of associated stimuli. Instead they performed a 1-back task on the currently viewed images that was in- dependent of any recall or associative learning. The changes in fMRI representations were observed while participants performed this 1-back task and in the pres- information ence of competing visual stimuli (e.g., about chair stimuli could be decoded while participants were actually viewing and performing a task on faces). Thus, although we cannot discount the possibility that participants automatically recalled a chair while viewing the associated face, this recall must necessarily have oc- curred in the presence of competing visual input and simultaneously with the performance of a nontrivial, in- dependent task performed on the currently perceived stimuli and that did not require explicit recall. In the end, as discussed above, although such automatic recall could be one of the possible mechanisms underlying the increase in cross-classification performance in our experiment, it is still consistent with the conclusion that multivoxel object representations can be flexibly modi- fied through associative learning. The finding that the largest learning-dependent changes (>9% increase in cross-classification performance) were
observed in clusters of voxels in the left and right fusiform
gyri is consistent with a previous study showing associative
learning effects in the left fusiform cortex (Park, Shannon,
Biggan, & Spann, 2012). The voxels showing the largest
changes overlapped substantially with our functionally
defined FFA, as well as with an anterior subdivision of
the LOC located in the fusiform gyrus (pF; Grill-Spector
et al., 1999). The object-selective pF itself partially overlap-
ped with the FFA (Grill-Spector et al., 2001), but we were
unable to further segregate these two ROIs in the native
space of each participant. Other recent studies have also
reported a mix of face- and object-selective voxels in the
traditionally defined FFA (Cukur, Huth, Nishimoto, &
Gallant, 2013; Hanson & Schmidt, 2011). It has been ar-
gued that increased expertise with a class of objects is cor-
related with the level of activation in the FFA (McGugin,
Gatenby, Gore, & Gauthier, 2012; Gauthier et al., 1999),
although this claim is still debated (McKone, Kanwisher, &
Duchaine, 2007; Kanwisher, 2000). Although our data are
unable to shed light on this debate because of the spatial
overlap between the FFA and the pF, we find that face- E
object-selective representations in the fusiform gyrus show
the strongest changes in representational similarity as a
result of associative learning.

Previous studies have investigated the effects of training
on object representations in object-selective cortex. In gen-
eral, these studies reveal that training-related changes occur
in a distributed fashion in inferotemporal cortex and that
these changes are often modest (Op de Beeck & Baker,
2010). In monkeys, training changes the selectivity and
strength of neuronal responses in inferotemporal cortex
(Li & DiCarlo, 2008; Freedman, Riesenhuber, Poggio, &
Mugnaio, 2006; Baker, Behrmann, & Olson, 2002; Sigala &
Logothetis, 2002; Logothetis, Pauls, & Poggio, 1995). Umano
fMRI studies have shown that learning is associated with in-
creases or decreases in the overall amplitude of the average
BOLD response (Op de Beeck, Baker, DiCarlo, & Kanwisher,
2006; Kourtzi, Betts, Sarkheil, & Welchman, 2005; Gauthier
et al., 1999), as well as with a sharpening of neural tuning
(Zhang, Meeson, Welchman, & Kourtzi, 2010; Gillebert,
Op de Beeck, Panis, & Wagemans, 2009; Jiang et al., 2007).
The current study extends this previous work by investigat-
ing the effects of associative learning on preexisting, BENE-
established response patterns for pairs of familiar categories
(rather than extensive practice with a single category).

864

Journal of Cognitive Neuroscience

Volume 28, Numero 6

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

Previous monkey studies have investigated a class of
neurons known as “pair-coding neurons” that respond
similarly to pairs of stimuli that have been associated to-
gether (Sakai & Miyashita, 1991). In these studies, mon-
keys learned associations between novel, meaningless
fractal patterns that they had been exposed to on a rela-
tively short timescale (cioè., in recent experimental ses-
sions). After learning, a neuron that was originally
selective to a cue stimulus showed selective responses
to its paired associate as well. Tuttavia, neuronal selec-
tivity for novel stimuli (per esempio., the cue stimuli in the afore-
mentioned studies) can flexibly develop as a result of
recent exposure (Logothetis et al., 1995), suggesting that
the pair-coding task principally modified neuronal re-
sponses in recently created representations. In contrasto,
our participants learned novel associations between al-
ready overlearned categories of stimuli, with which they
had lifelong exposure. After learning, we found that cat-
egory selectivity was modified in well-established (E
hence presumably less flexible) multivoxel representa-
tions that are thought to contribute to visual categoriza-
tion and object representation. Additionally, pair-coding
neurons show significantly correlated responses to pairs of
pictures (cioè., at the exemplar level) in a stimulus–stimulus
association task. In contrasto, we found that category level
multivoxel representations change, although the associa-
tions were created between exemplars of the two catego-
ries. Finalmente, pair-coding neurons have typically been
found in the anterior ventral portion of area TE and in
the perirhinal cortex (although a larger proportion of
these neurons and stronger pair-coding effects were
found in the perirhinal cortex; Naya, Yoshida, & Miyashita,
2003). Other studies have also found evidence for asso-
ciative learning in perirhinal cortex and anterior ventral
IT neurons in monkeys (Eifuku, Nakata, Sugimori, Ono,
& Tamura, 2010; Erickson & Desimone, 1999) and in sin-
gle neurons in the human MTL (Ison et al., 2015; Reddy
et al., 2015). Tuttavia, information about associated
stimuli has not been found in single neurons in more pos-
terior portions of TE (Gochin, Colombo, Dorfman,
Gerstein, & Gross, 1994). In this study, we observed the
strongest effects of associative learning in voxels in the fu-
siform cortex, overlapping with the FFA and pF. Although it
is difficult to establish exact homologies between the mon-
key and human brains, the human LOC and FFA are
thought to correspond to the posterior and dorsal part of
the monkey inferotemporal complex (Tsao, Moeller, &
Freiwald, 2008; Denys et al., 2004). Our findings thus sug-
gest that information about associated stimulus pairs is also
observed in human visual regions more caudal to those
previously reported in single neurons in monkey anterior
ventral inferotemporal cortex.

Acquiring new associations depends critically on MTL
structures, including the hippocampus (Squire, Stark, &
Clark, 2004; Fortin, Agster, & Eichenbaum, 2002). As
mentioned above, single-neuron recordings in monkeys
( Wirth et al., 2003; Erickson & Desimone, 1999; Sakai

& Miyashita, 1991; Miyashita, 1988) and humans (Ison
et al., 2015; Reddy et al., 2015) show that MTL neurons
change their selectivity as a result of learning associa-
tions between pairs of stimuli. Human fMRI studies have
implicated different MTL structures in associative learn-
ing, sequence learning, and relational memory (Schapiro,
Kustner, & Turk-Browne, 2012; Turk-Browne, Scholl,
Chun, & Johnson, 2009; Haskins, Yonelinas, Quamme, &
Ranganath, 2008; Aminoff, Gronau, & Bar, 2007; Diana,
Yonelinas, & Ranganath, 2007; Davachi, 2006; Prince,
Daselaar, & Cabeza, 2005). In particular, hippocampal fMRI
activity patterns become more similar to each other as a
result of incidental sequence learning (Schapiro et al.,
2012). In our study however, object category multivoxel
representations in the hippocampus were essentially
unmodified during the postlearning scan. This difference
between the two studies could be accounted for by differ-
ences in the learning protocols. For instance, in the pre-
vious study participants viewed sequences of items but
were unaware of the relationships between them. In our
study however, participants were explicitly instructed to
make associations between the object categories. Addition-
alleato, in our study learning occurred over a much longer
time frame, with the result that the associations were over-
learned (Figura 3) when postlearning brain activity was
measured. Così, although the hippocampus undoubtedly
plays an active role during the acquisition of new associa-
zioni, for instance by differentially activating for success-
fully learned versus unlearned associations (Davachi,
2006), it is possible that the relevant information was proc-
essed and stored in other cortical areas once the associa-
tions were overlearned. Infatti, although it is not known
how long memory traces need to remain active in MTL
structures before being committed to long-term storage
in anterior inferotemporal cortex, the representational
changes we observe in the VVP could be consistent with
such a reorganization of learned information.

Participants were explicitly asked to learn arbitrary as-
sociations between unrelated object categories, and we
measured changes in neural response patterns in an fMRI
scan session at the end of learning. Learned associations
in this case could be direct and automatic or mediated by
explicit strategies such as recall (as described earlier)
and/or visual imagery. Visual perception and visual imag-
ery of familiar categories of objects have been shown to
elicit similar patterns of fMRI activity in ventral visual cor-
tex (Reddy, Tsuchiya, & Serre, 2010). Recall of past visual
stimuli also reactivates their representations in visual cor-
tex ( Wheeler et al., 2000). It is conceivable that during
the postlearning scan of the current study, while viewing
one category of images (per esempio., chairs), participants brought
the associated category (per esempio., faces) to mind, although
they performed a 1-back task on the images that was inde-
pendent of any associative learning. Tuttavia, note that
even if participants could not avoid recall and mental im-
agery of the associated categories, the very experience of a
stimulus “bringing another to mind” when the task (1-back)

Senoussi et al.

865

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

does not require such recall is arguably a manifestation of
a well-learned association.

To conclude, we show that associative learning is ac-
companied by large-scale neural changes in the VVP. Spe-
cifically, multivoxel activity patterns for associated object
categories become more similar to each other with learn-
ing. An interesting open question that we have not ad-
dressed here is whether these representational changes
are specific to the stimuli with which learning occurred,
or whether they generalize to other exemplars in the cat-
egory. Additionally, how long do these changes persist
after the learned associations are no longer behaviorally
relevant? Although these questions remain exciting topics
for future research, here we show evidence for flexible
and dynamic representations in ventral temporal cortex
that could support the daily process of learning new re-
lationships between different events.

Ringraziamenti

This work was supported by funding from an ANR-JCJC (2012)
to L. R. and funding from the Institute des Sciences du Cerveau
de Toulouse to L. R. and R. V. We thank Francisco Pereira for
sharing his SearchMight toolbox with us. l. R. and R. V. Di-
signed the research. The authors would like to thank the staff
of the Imaging Center, INSERM/UPS UMR 825 MRI platform for
their assistance in acquiring the data.

Reprint requests should be sent to Dr. Leila Reddy, CNRS-Centre
de Recherche Cerveau et Cognition, Pavillon Baudot CHU
Purpan, 31052 Toulouse Cedex, France, or via e-mail: leila.reddy@
cerco.ups-tlse.fr.

REFERENCES

Aminoff, E., Gronau, N., & Bar, M. (2007). The parahippocampal
cortex mediates spatial and nonspatial associations. Cerebral
Cortex, 17, 1493–1503.

Baker, C. I., Behrmann, M., & Olson, C. R. (2002). Impact
of learning on representation of parts and wholes in
monkey inferotemporal cortex. Nature Neuroscience,
5, 1210–1216.

Carlson, T. A., Schrater, P., & Lui, S. (2003). Patterns of activity
in the categorical representations of objects. Journal of
Cognitive Neuroscience, 15, 704–717.

Cukur, T., Huth, UN. G., Nishimoto, S., & Gallant, J. l. (2013).
Functional subdomains within human FFA. Journal of
Neuroscience, 33, 16748–16766.

Davachi, l. (2006). Item, context and relational episodic

encoding in humans. Current Opinion in Neurobiology,
16, 693–700.

Denys, K., Vanduffel, W., Fize, D., Nelissen, K., Peuskens, H.,

Van Essen, D., et al. (2004). The processing of visual shape in
the cerebral cortex of human and nonhuman primates: UN
functional magnetic resonance imaging study. Journal of
Neuroscience, 24, 2551–2565.

Diana, R. A., Yonelinas, UN. P., & Ranganath, C. (2007). Imaging
recollection and familiarity in the medial temporal lobe: UN
three-component model. Trends in Cognitive Sciences,
11, 379–386.

Downing, P. E., Jiang, Y., Shuman, M., & Kanwisher, N. (2001).
A cortical area selective for visual processing of the human
body. Scienza, 293, 2470–2473.

Eifuku, S., Nakata, R., Sugimori, M., Ono, T., & Tamura, R.

(2010). Neural correlates of associative face memory in the
anterior inferior temporal cortex of monkeys. Journal of
Neuroscience, 30, 15085–15096.

Epstein, R., & Kanwisher, N. (1998). A cortical representation of

the local visual environment. Nature, 392, 598–601.

Erickson, C. A., & Desimone, R. (1999). Responses of macaque
perirhinal neurons during and after visual stimulus association
apprendimento. Journal of Neuroscience, 19, 10404–10416.

Fortin, N. J., Agster, K. L., & Eichenbaum, H. B. (2002). Critical
role of the hippocampus in memory for sequences of events.
Nature Neuroscience, 5, 458–462.

Freedman, D. J., Riesenhuber, M., Poggio, T., & Mugnaio, E. K.
(2006). Experience-dependent sharpening of visual shape
selectivity in inferior temporal cortex. Cerebral Cortex,
16, 1631–1644.

Gauthier, I., Tarr, M. J., Anderson, UN. W., Skudlarski, P., & Gore,
J. C. (1999). Activation of the middle fusiform ‘face area’
increases with expertise in recognizing novel objects.
Nature Neuroscience, 2, 568–573.

Gillebert, C. R., Op de Beeck, H. P., Panis, S., & Wagemans, J.
(2009). Subordinate categorization enhances the neural
selectivity in human object-selective cortex for fine shape
differences. Journal of Cognitive Neuroscience, 21,
1054–1064.

Gochin, P. M., Colombo, M., Dorfman, G. A., Gerstein, G. L.,

& Gross, C. G. (1994). Neural ensemble coding in
inferior temporal cortex. Journal of Neurophysiology,
71, 2325–2337.

Gordon, UN. M., Rissman, J., Kiani, R., & Wagner, UN. D. (2014).
Cortical reinstatement mediates the relationship between
content-specific encoding activity and subsequent
recollection decisions. Cerebral Cortex, 24, 3350–3364.
Grill-Spector, K. (2003). The neural basis of object perception.

Current Opinion in Neurobiology, 13, 159–166.

Grill-Spector, K., Kourtzi, Z., & Kanwisher, N. (2001). The lateral
occipital complex and its role in object recognition. Vision
Research, 41, 1409–1422.

Grill-Spector, K., Kushnir, T., Edelman, S., Avidan, G., Itzchak,
Y., & Malach, R. (1999). Differential processing of objects
under various viewing conditions in the human lateral
occipital complex. Neuron, 24, 187–203.

Grill-Spector, K., & Malach, R. (2004). The human visual cortex.

Annual Review of Neuroscience, 27, 649–677.

Hanke, M., Halchenko, Y. O., Sederberg, P. B., Hanson, S. J.,
Haxby, J. V., & Pollmann, S. (2009). PyMVPA: A python
toolbox for multivariate pattern analysis of fMRI data.
Neuroinformatics, 7, 37–53.

Hanson, S. J., & Schmidt, UN. (2011). High-resolution imaging of
the fusiform face area (FFA) using multivariate non-linear
classifiers shows diagnosticity for non-face categories.
Neuroimage, 54, 1715–1734.

Harel, A., Kravitz, D. J., & Baker, C. IO. (2014). Task context
impacts visual object processing differentially across the
cortex. Proceedings of the National Academy of Sciences,
U.S.A., 111, E962–E971.

Haskins, UN. L., Yonelinas, UN. P., Quamme, J. R., & Ranganath,

C. (2008). Perirhinal cortex supports encoding and
familiarity-based recognition of novel associations.
Neuron, 59, 554–560.

Haxby, J. V., Gobbini, M. I., Furey, M. L., Ishai, A., Schouten,
J. L., & Pietrini, P. (2001). Distributed and overlapping
representations of faces and objects in ventral temporal
cortex. Scienza, 293, 2425–2430.

Haxby, J. V., Guntupalli, J. S., Connolly, UN. C., Halchenko, Y. O.,

Conroy, B. R., Gobbini, M. I., et al. (2011). A common,
high-dimensional model of the representational space in
human ventral temporal cortex. Neuron, 72, 404–416.

866

Journal of Cognitive Neuroscience

Volume 28, Numero 6

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

Haynes, J. D., & Rees, G. (2005). Predicting the orientation of
invisible stimuli from activity in human primary visual cortex.
Nature Neuroscience, 8, 686–691.

Miyashita, Y., & Chang, H. S. (1988). Neuronal correlate of

pictorial short-term memory in the primate temporal cortex.
Nature, 331, 68–70.

Ison, M. J., Quian Quiroga, R., & Fried, IO. (2015). Rapid

Naya, Y., Yoshida, M., & Miyashita, Y. (2003). Forward

encoding of new memories by individual neurons in the
human brain. Neuron, 87, 220–230.

Jiang, X., Bradley, E., Rini, R. A., Zeffiro, T., Vanmeter, J., &
Riesenhuber, M. (2007). Categorization training results in
shape- and category-selective human neural plasticity.
Neuron, 53, 891–903.

Johnson, J. D., McDuff, S. G., Rugg, M. D., & Norman, K. UN.

(2009). Recollection, familiarity, and cortical reinstatement:
A multivoxel pattern analysis. Neuron, 63, 697–708.
Kamitani, Y., & Tong, F. (2005). Decoding the visual and

subjective contents of the human brain. Nature
Neuroscience, 8, 679–685.

Kanwisher, N. (2000). Domain specificity in face perception.

Nature Neuroscience, 3, 759–763.

processing of long-term associative memory in monkey
inferotemporal cortex. Journal of Neuroscience, 23,
2861–2871.

Nyberg, L., Habib, R., McIntosh, UN. R., & Tulving, E. (2000).
Reactivation of encoding-related brain activity during
memory retrieval. Proceedings of the National Academy
of Sciences, U.S.A., 97, 11120–11124.

Op de Beeck, H. P., & Baker, C. IO. (2010). The neural basis

of visual object learning. Trends in Cognitive Sciences, 14,
22–30.

Op de Beeck, H. P., Baker, C. I., DiCarlo, J. J., & Kanwisher,

N. G. (2006). Discrimination training alters object
representations in human extrastriate cortex. Journal of
Neuroscience, 26, 13025–13036.

Kanwisher, N., McDermott, J., & Chun, M. M. (1997). IL

Op de Beeck, H. P., Brants, M., Baeck, A., & Wagemans, J.

fusiform face area: A module in human extrastriate cortex
specialized for face perception. Journal of Neuroscience,
17, 4302–4311.

Kourtzi, Z., Betts, l. R., Sarkheil, P., & Welchman, UN. E. (2005).
Distributed neural plasticity for shape learning in the human
visual cortex. PLoS Biology, 3, e204.

Kriegeskorte, N., Goebel, R., & Bandettini, P. (2006).

Information-based functional brain mapping. Proceedings of
the National Academy of Sciences, U.S.A., 103, 3863–3868.

Kriegeskorte, N., Mur, M., & Bandettini, P. (2008).

Representational similarity analysis—Connecting the
branches of systems neuroscience. Frontiers in Systems
Neuroscience, 2, 4.

Kriegeskorte, N., Mur, M., Ruff, D. A., Kiani, R., Bodurka, J.,

Esteky, H., et al. (2008). Matching categorical object
representations in inferior temporal cortex of man and
monkey. Neuron, 60, 1126–1141.

Kriegeskorte, N., Simmons, W. K., Bellgowan, P. S., & Baker,
C. IO. (2009). Circular analysis in systems neuroscience:
The dangers of double dipping. Nature Neuroscience,
12, 535–540.

Kuhl, B. A., & Chun, M. M. (2014). Successful remembering

elicits event-specific activity patterns in lateral parietal cortex.
Journal of Neuroscience, 34, 8051–8060.

Kuhl, B. A., Rissman, J., Chun, M. M., & Wagner, UN. D. (2011).
Fidelity of neural reactivation reveals competition between
ricordi. Proceedings of the National Academy of
Scienze, U.S.A., 108, 5903–5908.

Li, N., & DiCarlo, J. J. (2008). Unsupervised natural experience
rapidly alters invariant object representation in visual cortex.
Scienza, 321, 1502–1507.

Logothetis, N. K., Pauls, J., & Poggio, T. (1995). Shape

representation in the inferior temporal cortex of monkeys.
Current Biology, 5, 552–563.

McGugin, R. W., Gatenby, J. C., Gore, J. C., & Gauthier, IO.

(2012). High-resolution imaging of expertise reveals reliable
object selectivity in the fusiform face area related to
perceptual performance. Proceedings of the National
Academy of Sciences, U.S.A., 109, 17063–17068.

(2010). Distributed subordinate specificity for bodies, faces,
and buildings in human ventral visual cortex. Neuroimage,
49, 3414–3425.

Park, H., Shannon, V., Biggan, J., & Spann, C. (2012). Neural
activity supporting the formation of associative memory
versus source memory. Brain Research, 1471, 81–92.

Polyn, S. M., Natu, V. S., Cohen, J. D., & Norman, K. UN. (2005).
Category-specific cortical activity precedes retrieval during
memory search. Scienza, 310, 1963–1966.

Prince, S. E., Daselaar, S. M., & Cabeza, R. (2005). Neural

correlates of relational memory: Successful encoding and
retrieval of semantic and perceptual associations. Journal
of Neuroscience, 25, 1203–1210.

Reddy, L., & Kanwisher, N. (2006). Coding of visual objects
in the ventral stream. Current Opinion in Neurobiology,
16, 408–414.

Reddy, L., & Kanwisher, N. (2007). Category selectivity in the
ventral visual pathway confers robustness to clutter and
diverted attention. Current Biology, 17, 2067–2072.

Reddy, L., Poncet, M., Self, M. W., Peters, J. C., Douw, L., van
Dellen, E., et al. (2015). Learning of anticipatory responses in
single neurons of the human medial temporal lobe. Nature
Communications, 6, 8556.

Reddy, L., Tsuchiya, N., & Serre, T. (2010). Reading the mind’s
eye: Decoding category information during mental imagery.
Neuroimage, 50, 818–825.

Sakai, K., & Miyashita, Y. (1991). Neural organization for the

long-term memory of paired associates. Nature, 354,
152–155.

Schapiro, UN. C., Kustner, l. V., & Turk-Browne, N. B. (2012).
Shaping of object representations in the human medial
temporal lobe based on temporal regularities. Current
Biology, 22, 1622–1627.

Sigala, N., & Logothetis, N. K. (2002). Visual categorization
shapes feature selectivity in the primate temporal cortex.
Nature, 415, 318–320.

Spiridon, M., & Kanwisher, N. (2002). How distributed is visual
category information in human occipito-temporal cortex?
An fMRI study. Neuron, 35, 1157–1165.

McKone, E., Kanwisher, N., & Duchaine, B. C. (2007). Can

Squire, l. R., Stark, C. E., & Clark, R. E. (2004). The medial

generic expertise explain special processing for faces? Trends
in Cognitive Sciences, 11, 8–15.

Messinger, A., Squire, l. R., Zola, S. M., & Albright, T. D. (2001).
Neuronal representations of stimulus associations develop
in the temporal lobe during learning. Atti del
National Academy of Sciences, U.S.A., 98, 12239–12244.
Miyashita, Y. (1988). Neuronal correlate of visual associative

long-term memory in the primate temporal cortex. Nature,
335, 817–820.

temporal lobe. Annual Review of Neuroscience,
27, 279–306.

Stansbury, D. E., Naselaris, T., & Gallant, J. l. (2013). Natural
scene statistics account for the representation of scene
categories in human visual cortex. Neuron,
79, 1025–1034.

Straw, UN. D. (2008). Vision egg: An open-source library for

realtime visual stimulus generation. Frontiers in
Neuroinformatics, 2, 4.

Senoussi et al.

867

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

Tsao, D. Y., Moeller, S., & Freiwald, W. UN. (2008). Comparing
face patch systems in macaques and humans. Proceedings
of the National Academy of Sciences, U.S.A.,
105, 19514–19519.

Turk-Browne, N. B., Scholl, B. J., Chun, M. M., & Johnson, M. K.

(2009). Neural evidence of statistical learning: Efficient
detection of visual regularities without awareness. Journal of
Cognitive Neuroscience, 21, 1934–1945.

Wheeler, M. E., Petersen, S. E., & Buckner, R. l. (2000). Memory’s
echo: Vivid remembering reactivates sensory-specific cortex.

Proceedings of the National Academy of Sciences, U.S.A.,
97, 11125–11129.

Wirth, S., Yanike, M., Frank, l. M., Smith, UN. C., Brown, E. N.,
& Suzuki, W. UN. (2003). Single neurons in the monkey
hippocampus and learning of new associations. Scienza,
300, 1578–1581.

Zhang, J., Meeson, A., Welchman, UN. E., & Kourtzi, Z. (2010).
Learning alters the tuning of functional magnetic resonance
imaging patterns for visual forms. Journal of Neuroscience,
30, 14127–14133.

l

D
o
w
N
o
UN
D
e
D

F
R
o
M
H

T
T

P

:
/
/

D
io
R
e
C
T
.

M

io
T
.

e
D
tu

/
j

/

o
C
N
UN
R
T
io
C
e
–
P
D

l

F
/

/

/

/

2
8
6
8
5
2
1
7
8
4
9
5
7

/
j

o
C
N
_
UN
_
0
0
9
3
3
P
D

.

F

B
sì
G
tu
e
S
T

T

o
N
0
8
S
e
P
e
M
B
e
R
2
0
2
3

868

Journal of Cognitive Neuroscience

Volume 28, Numero 6

Scarica il pdf