10⁹ into 109. The quotes are reproduced exactly as the extraction left them and have NOT been repaired — the meaning of such a code point belongs to the one PDF it came from, and guessing it would produce a clean, confident, unverifiable quote.phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | short-term memory localisation in mouse neocortex— “The location of short-term memory in mammalian neocortex remains elusive.” | |
| claim-type | mechanism— “in mammalian neocortex remains elusive. Here we show that distinct neocortical areas maintain short-term memory depending on behavioral strategy. Using wide-field and single-cell calcium imaging,” | |
| method — evidence | — “Using wide-field and single-cell calcium imaging, we measured layer 2/3 neuronal activity in mice performing a whisker-based texture discrimination task with delayed response.” | |
| mechanism | Posterior parietal area RL, which projects both frontally to M2 and posteriorly to area P, routes the sensory information to one destination or the other according to the animal's behavioural state— “Therefore, they could route information to either frontal or posterior cortices depending on the mouse’s state.” | mechanisms[0] |
| mechanism | When information is held frontally in M2, posterior cortex is actively suppressed below baseline, which is why the two memory patterns are mutually exclusive— “The latter observation suggests that posterior areas may be inhibited when information is maintained in M2 (Allen et al., 2017).” | mechanisms[1] |
| replicates | Chen et al., 2016; Yang et al., 2015 (choice encoding in posterior sensory areas)— “Thus, consistent with previous reports (Chen et al., 2016; Yang et al., 2015), sensory-related posterior areas S1BC, S2 and RL encode choice shortly after touch for both behavioral strategies.” | claim_relations[0] |
| replicates | Chen et al., 2013, 2016 (timing of hit/CR discrimination)— “Discrimination between hit and CR trials occurred within a few hundred milliseconds, in line with previous studies (Chen et al., 2013, 2016).” | claim_relations[1] |
| extends | Guo et al., 2014; Li et al., 2015 (frontal areas including ALM as the site of short-term memory)— “In passive trials, M2 as well as other frontal areas (including ALM; Guo et al., 2014; Li et al., 2015) were not activated and could not discriminate trial types (Figure 4), suggesting that they are less important for maintaining information for this type of behavior, at least before any preparatory movements.” | claim_relations[2] |
| builds-on | Leavitt et al., 2017 (sustained activity encoding working memory is not fully distributed)— “Maintenance of information in short-term memory is not necessarily tied to one cortical area but is flexibly localized according to behavioral strategy (Leavitt et al., 2017).” | claim_relations[3] |
| sense | location of short-term memory - NOT A SINGLE SITE: medio-frontal on active trials and lateral posterior on passive ones, so the answer is conditional on the animal's strategy— “We propose that short-term memory is present in two different forms: motor-related memory in M2 or sensory-related memory in P” | senses[0] |
| sense | behavioral strategy - an experimental grouping made from the animal's own conduct: engaging the body toward the approaching texture, or passively awaiting the touch— “We have shown that mice can deploy either an active or passive strategy to solve a texture discrimination task.” | senses[1] |
| sense | active versus passive trial - operationalised by thresholding body movement during the sensation period, so the class is fixed by a cut on a continuous measure— “Classification into active (dark green) and passive (light green) trials is based on movement during the sensation period.” | senses[2] |
| sense | delay period - AS ANALYSED, NOT AS RUN: restricted to `quiet trials` in which the mouse did not move for at least the first second, with traces truncated at the first movement thereafter. THIS IS A CONTROL, NOT A CONFOUND - the classes are defined by movement during the SENSATION period, and the delay analysis is made movement-free on purpose so that the persistent activity cannot be motor activity— “To analyze the delay period, we focused only on quiet trials where the mouse did not move for at least the first second in the delay (starting 0.2 s after beginning of texture retraction; Figure 3A). Trials where the mouse moved during the first second of the delay period where defined as noisy and discarded from the analysis (Figure 1D… | senses[3] |
| control | a set of control experiments on the imaging signal itself — held — rules out autofluorescence and non-calcium-related intrinsic signals as the source of the measured activity— “In control experiments we excluded confounding effects of autofluorescence or non-calcium-related intrinsic signals (STAR Methods and Figure S3).” | controls[0] |
| control | control imaging targeted at the near-midline M2 area, which lies close to the superior sagittal sinus — held — rules out hemodynamic signal as the source of the activity patterns, in the one location where a blood vessel makes it most likely— “In control experiments, we verified that these activity patterns were not confounded by hemodynamic signals, especially for the near-midline M2 area, which is close to the superior sagittal sinus (Figures S3 and S8).” | controls[1] |
| control | training two wild-type mice - carrying no GCaMP - on the same task and imaging them awake and performing — held — rules out autofluorescence, by running the whole experiment in animals that cannot produce the signal of interest— “2) To control for autofluorescence signals we trained 2 wild-type mice on the task and obtained wide-field imaging in awake performing animals (Figure S3A-D).” | controls[2] |
| control | cellular-resolution imaging run alongside the wide-field measurement — held — rules out the possibility that the bulk population signal is not driven by the layer 2/3 cell bodies it is interpreted as reflecting— “In summary, the results from all the above control experiments indicate no major influence of non-calcium-related intrinsic signals on the GCaMP6f signals that would confound our signal interpretations.” | controls[3] |
| control | four experiments on the light delivery, including shining light on cortex not expressing ArchT in the same mice - one record for the set, quoting the sentence that reports the outcome — held — rules out light disturbance itself as the cause of the behavioural effect, and establishes that the effect requires the right population at the right time— “In summary, these control experiments demonstrate that light delivered above the wide-field preparation alone does not affect the performance of the mice but that it needs to reach the relevant neuronal populations at the right time to become effective.” | controls[4] |
| outcome (stated by the tagger, no quotation) | control — held | controls[0] |
| outcome (stated by the tagger, no quotation) | control — held | controls[1] |
| outcome (stated by the tagger, no quotation) | control — held | controls[2] |
| outcome (stated by the tagger, no quotation) | control — held | controls[3] |
| outcome (stated by the tagger, no quotation) | control — held | controls[4] |
| for which use | limitation — every signal reported: WIDE-FIELD FLUORESCENCE SIGNALS REPRESENT BULK SIGNALS, SUMMED OVER NEURONAL POPULATIONS (AS WELL AS POTENTIALLY CONFOUNDED BY NON-NEURONAL FLUORESCENCE CHANGES, WHICH NEEDS TO BE CONTROLLED FOR)— “The disadvantage of the method is that the wide-field fluorescence signals represent bulk signals, summed over neuronal populations (as well as potentially confounded by non-neuronal fluorescence changes, which needs to be controlled for).” | limitations[0] |
| for which use | limitation — the laminar reach of the conclusions: THIS STUDY FOCUSED ON LAYER 2/3 NEURONS, WHEREAS OTHER LAYERS MAY CONTRIBUTE DIFFERENTLY TO TASK-RELATED PROCESSES— “In addition, this study focused on layer 2/3 neurons, whereas other layers may contribute differently to task-related processes (Chandrasekaran et al., 2017; van Kerkoerle et al., 2014; Li et al., 2015).” | limitations[1] |
| for which use | limitation — the anatomical identity of the posterior activation: GIVEN THE VARIABILITY AMONG ANIMALS ... THE OBSERVED ACTIVATION SPOT MAY ALSO ENCOMPASS ADJACENT POSTERIOR AREAS SUCH AS LATEROMEDIAL VISUAL CORTEX OR POSTRHINAL CORTICES— “Given the variability among animals, however, the observed activation spot may also encompass adjacent posterior areas such as lateromedial visual cortex or postrhinal cortices.” | limitations[2] |
| for which use | limitation — the what/where pathway interpretation: ADDITIONAL EXPERIMENTS DIRECTLY TESTING EACH PATHWAY WILL BE NEEDED TO SUBSTANTIATE THESE IDEAS— “Nevertheless, additional experiments directly testing each pathway will be needed to substantiate these ideas.” | limitations[3] |
| for which use | limitation — the absence of task-relevant activity in an area of these maps: the signal is a bulk population measure and the authors state intermingled subpopulations could be averaged out within it— “wide-field fluorescence signals represent bulk signals, summed over neuronal populations (as well as potentially confounded by non-neuronal fluorescence changes, which needs to be controlled for). We therefore cannot exclude the possibility that we have missed intermingled sub- 28 populations of neurons that carry task-relevant information… | limitations[4] |
| for which use | limitation — the active/passive trial labels in this study: in sessions without usable body-camera images they were predicted from a z-axis focus drop, reported at 75% accuracy against the camera— “For a small subset of sessions where the body camera gave low quality images or was missing, we predicted the mouse’s movement based on the z-axis drop in focus (using a 10% drop from baseline in 3 in-focus single cells as criterion; 75% accuracy in labeling trials by validating with move vector from the body camera).” | limitations[5] |
| replication | untested | |
| limitation | Wide-field signals are bulk population signals that may be confounded by non-neuronal fluorescence— “The disadvantage of the method is that the wide-field fluorescence signals represent bulk signals, summed over neuronal populations (as well as potentially confounded by non-neuronal fluorescence changes, which needs to be controlled for).” | limitations[0] |
| limitation | Only layer 2/3 was studied; other layers may behave differently— “In addition, this study focused on layer 2/3 neurons, whereas other layers may contribute differently to task-related processes (Chandrasekaran et al., 2017; van Kerkoerle et al., 2014; Li et al., 2015).” | limitations[1] |
| limitation | The identity of the posterior activation spot is uncertain and may include neighbouring areas— “Given the variability among animals, however, the observed activation spot may also encompass adjacent posterior areas such as lateromedial visual cortex or postrhinal cortices.” | limitations[2] |
| limitation | The proposed what/where pathway interpretation is not directly tested here— “Nevertheless, additional experiments directly testing each pathway will be needed to substantiate these ideas.” | limitations[3] |
| limitation | The measurement is a bulk signal, and the authors state they cannot exclude having missed intermingled subpopulations averaged out within it.— “wide-field fluorescence signals represent bulk signals, summed over neuronal populations (as well as potentially confounded by non-neuronal fluorescence changes, which needs to be controlled for). We therefore cannot exclude the possibility that we have missed intermingled sub- 28 populations of neurons that carry task-relevant information… | limitations[4] |
| limitation | The grouping variable the entire result rests on is, in some sessions, assigned by a substitute measure whose accuracy is stated as 75%.— “For a small subset of sessions where the body camera gave low quality images or was missing, we predicted the mouse’s movement based on the z-axis drop in focus (using a 10% drop from baseline in 3 in-focus single cells as criterion; 75% accuracy in labeling trials by validating with move vector from the body camera).” | limitations[5] |
| limitation | The independent variable is changed by the manipulation. Strategy is treated as what determines where short-term memory sits, but optogenetic silencing of the site induces strategy switches that persist across days and appear in non-perturbation trials. The paper reports this as a result rather than as a threat, and it is the correct reading - but it means strategy and site are mutually determining, not one prior to the other, and the title states only one direction.— “Here, we find that some mice manage to maintain relatively high performance despite perturbation of one area (M2 or P) whereas other mice show significantly reduced performance. The former mice apparently flexibly switch to the alternative unperturbed strategy whereas the latter mice maintain their strategy. Strategy switches were consistent… | limitations[6] |
| limitation | Two thresholds on the same movement signal do different jobs: one splits trials into active and passive by movement during the sensation period, the other discards trials with movement in the first second of the delay as noisy. The grouping rule and the exclusion rule read the same measure, so which trials survive is not independent of which class they were assigned to.— “Dashed black line indicates threshold to construct the corresponding binary movement vectors (lower black traces). Classification into active (dark green) and passive (light green) trials is based on movement during the sensation period. Trials with the mouse moving during the first second in the delay period were defined as noisy and… | limitations[7] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | Which cortical area held the delay-period memory depended on the mouse's behavioural strategy: persistent activity sat medio-frontally (M2) on active trials but in a lateral posterior area on passive trials.— “During the delay, in contrast, persistent cortical activity was located medio-frontally in active trials but in a lateral posterior area in passive trials.” | content_tags.findings[0] |
| finding | The posterior whisker areas encoded the animal's choice soon after touch under both strategies, so the strategy difference appeared only later, during the delay.— “texture-or passively awaited the touch. Independent of strategy, whisker-related posterior areas encoded choice early after touch. During the delay, in contrast,” | content_tags.findings[1] |
| finding | Optogenetically silencing M2 or area P lowered task performance only in trials run with the strategy that area supports, and left the other strategy unaffected.— “In summary, perturbation of one area (M2 or P) decreased performance exclusively when the mouse used the associated strategy (Figure 7D; P < 0.05; signed-rank test).” | content_tags.findings[2] |
| finding | Silencing M2 and area P together significantly reduced performance in all four VGAT-ChR2 mice, to levels not significantly different from the expert threshold.— “All four mice significantly reduced their performance upon ChR2 activation (P < 0.001; signed-rank test), down to levels not significantly different from the expert threshold (Figure 7F; P > 0.05; signed-rank test).” | content_tags.findings[3] |
| finding | One passive mouse switched its strategy from passive to active in the perturbation sessions.— “Notably, however, this mouse very efficiently switched its strategy from passive to active in the perturbation sessions (Figure 6Fii, iii; P < 0.001; Mann-Whitney U-test).” | content_tags.findings[4] |
| finding | The two memory sites hold different content: the motor-sensory index was positive in M2 (activity grouped by the upcoming action) and negative in area P (activity grouped by the past texture).— “MSI was significantly positive in M2 and significantly negative in P (Figure 8D; P < 0.05; Wilcoxon signed-rank test), implying that each area encodes a different type of short-term memory” | content_tags.findings[5] |
| finding | On passive trials a classifier restricted to frontal-cortex pixels was at chance distinguishing hit from correct-rejection trials.— “Interestingly, for passive trials, SVM accuracy was at chance level when we applied the SVM only to pixels in the frontal cortex (Figure 4F; P > 0.05; signed-rank test compared to 50%; Bonferroni corrected; n = 39 imaging sessions from 5 passive mice).” | content_tags.findings[6] |
| finding | Discrimination performance (d') correlated positively with how actively a mouse used its whiskers (r = 0.6).— “In addition, d’ performance positively correlated with activeness (Figure 1H; r = 0.6; P < 0.001)” | content_tags.findings[7] |
| quantity | 16 and 82% — range of activeness (percentage of hit trials with body movement at texture touch) across the eight imaged mice— “‘Activeness’, quantified as the percentage of hit trials with body movements prior and during the texture touch, varied across mice between 16 and 82%” | content_tags.quantities[0] |
| quantity | 60-88% — accuracy of the classifier separating hit from correct-rejection delay maps, across mice and both strategies— “accuracy ranged from 60-88% in all mice for active and passive trials (Figure 4B).” | content_tags.quantities[1] |
| quantity | 48.5 ± 4.2 and 50.4 ± 3.8 — classifier accuracy after shuffling trial labels, i.e. the chance baseline for active and passive trials— “To test for significance, we shuffled between hit and CR labels (while maintaining active and passive labels) and performed the same SVM, resulting in an accuracy of 48.5 ± 4.2 and 50.4 ± 3.8 (mean ± s.e.m), i.e. chance level, for active and passive trials, respectively.” | content_tags.quantities[2] |
| quantity | r = 0.6 — correlation between a session's activeness and the mouse's d-prime discrimination performance— “In addition, d’ performance positively correlated with activeness (Figure 1H; r = 0.6; P < 0.001)” | content_tags.quantities[3] |
| quantity | 6.9 ± 0.3 weeks — training time needed before imaging, across the eight mice— “During training, an increasing delay was gradually introduced over weeks (Figure 1B; training time 6.9 ± 0.3 weeks; mean ± s.e.m.; n = 8 mice).” | content_tags.quantities[4] |
| quantity | 1.2-4.5 s — range of enforced delay durations mice could withhold licking for— “Mice learned to withhold licking for up to several seconds (range of 1.2-4.5 s).” | content_tags.quantities[5] |
| quantity | 2.89° ± 0.36° vs. 0.95° ± 0.14° — whisking envelope around texture arrival on active versus passive trials— “mean whisking envelope from -1 to +1 seconds relative to texture stop: 2.89° ± 0.36° vs. 0.95° ± 0.14°; P < 0.05” | content_tags.quantities[6] |
| quantity | 30 — total adult male mice used across imaging, control and optogenetic experiments— “A total of 30 adult male mice (1-9 months old) were used in this study.” | content_tags.quantities[7] |
| variable | layer 2/3 neuronal activity (dependent)— “Using wide-field and single-cell calcium imaging we measured layer 2/3 neuronal activity in mice performing a whisker-based texture discrimination task with delayed response.” | content_tags.variables[2] |
| variable | d-prime (dependent)— “Performance was quantified as d-prime (Chen et al., 2013)” | content_tags.variables[3] |
| variable | body movement (control)— “Therefore, the data analysis for the delay period excluded any movement-related activity and reflects the population dynamics when the mouse is sitting quietly and maintaining information to guide its future action.” | content_tags.variables[4] |
| definition | activeness— “‘Activeness’ was defined as the percentage of active hit trials and derived for each imaging session” | content_tags.definitions[0] |
| definition | active versus passive trial— “A trial was labeled (unrelated to hit or CR) as ‘active’ if the mouse was moving for at least 0.5 seconds during the sensation period (-1 s to 2 s relative to texture stop) and ‘passive’ otherwise.” | content_tags.definitions[1] |
| definition | motor-sensory index (MSI)— “MSI ranges from -1 to 1 where positive values indicate greater encoding of motor-related information (similar activity for similar future action) whereas negative values indicate greater encoding of sensory-related information (similar activity for similar past stimuli).” | content_tags.definitions[2] |
| definition | posterior area P— “we refer to this region as ‘posterior area P’ (or brief ‘P’; see Discussion; see also additional retinotopic alignment in Figure S7B)” | content_tags.definitions[3] |
| concept | short-term memory— “We propose that short-term memory is present in two different forms: motor-related memory in M2 or sensory-related memory in P” | concepts[0] |
| concept | behavioral strategy— “We have shown that mice can deploy either an active or passive strategy to solve a texture discrimination task.” | concepts[1] |
| concept | wide-field— “wide-field fluorescence signals represent bulk signals, summed over neuronal populations” | concepts[2] |
| concept | activeness— “‘Activeness’, quantified as the percentage of hit trials with body movements prior and during the texture touch, varied across mice between 16 and 82% (Figure 1F; STAR Methods).” | concepts[3] |
| concept | quiet trials— “To analyze the delay period, we focused only on quiet trials where the mouse did not move for at least the first second in the delay (starting 0.2 s after beginning of texture retraction; Figure 3A). Trials where the mouse moved during the first second of the delay period where defined as noisy and discarded from the analysis (Figure 1D… | concepts[4] |
| concept | M2— “We propose that short-term memory is present in two different forms: motor-related memory in M2 or sensory-related memory… | concepts[5] |
| concept | z-axis drop in focus | concepts[6] |
| Wilcoxon signed-rank test and Mann-Whitney U-test P | as written: Mann-Whitney U-test ... Wilcoxon signed rank test | content_tags.statistics[0] |
| ANOVA F | as written: ANOVA and multiple group correction | content_tags.statistics[1] |
| F statistic with 95% confidence interval F | as written: F10,33 = 5.17, 95% confidence interval (CI) | content_tags.statistics[2] |
| Kolmogorov-Smirnov test P | as written: Kolmogorov-Smirnov test on residuals | content_tags.statistics[3] |
| support vector machine with cross-validation | as written: SVM classification ... 20 cross-validations | content_tags.statistics[4] |
| label-shuffling permutation test | as written: we shuffled between hit and CR labels | content_tags.statistics[5] |
| principal component analysis PCs | as written: principal component analysis (PCA) Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[6] |
| ROC analysis / area under the curve AUC | as written: receiver operating characteristics (ROC) curve ... area under the curve (AUC) Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[7] |
| correlation coefficient r | as written: positively correlated ... r = 0.6 Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[8] |
| Bonferroni correction P | as written: Bonferroni corrected Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[9] |
When we have changed our mind. Some entries in the table begin with a word in capitals and the reason for it. EVIDENCE CORRECTED — the finding still stands, but we had quoted the wrong sentence from the paper, and the quotation has been replaced with the right one. RETRACTED — the entry itself was wrong; it stays on the page, marked, with the reason, rather than being deleted. WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below. Each mark carries its own reason, in brackets, because the reasons differ from one entry to the next. An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason. A line reading Audit and a date is a later re-reading of the entry against its whole sentence: DOUBTFUL means it was kept, and the reason says what is loose about it. Quotation marks inside the tagger's sentence. Where a tag is the tagger's own sentence, a phrase in quotation marks inside it is borrowed from the paper — the tagger's wording around it is not. The paper's own sentence is the evidence quote for that row; on this page it is withheld with the article, and the article is at its publisher. Nothing inside quotation marks anywhere on this page is ours: if it is quoted, it is the paper's. Whose words a concept carries. A concept entry records a term the paper itself uses, so the term, the sense given for it, and its evidence are the paper's own words, quoted where the paper fixes them. This is the opposite of a sense entry in the tag table, which is the tagger's sentence about how the paper uses a term. Where a concept entry could not be grounded in the paper's words, the entry says so on its own row.
Tag layer © the Meta-Tagging Project, CC BY 4.0. The paper text is © its authors/publisher
and is linked, not reproduced. Excerpts shown are short quotations for identification and verification.
· back to index
· concepts · authors
· עברית
© 2026 Shir Sivroni · ORCID 0009-0000-2597-2824