Claims 45
From the paper Ventral pallidum neurons dynamically signal relative threat · claims file · checked 2026-11-06: ready
- recorded-units
435 VP neurons were recorded from 14 rats over 194 sessions.
A total of 435 VP neurons were recorded from 14 rats over 194 sessions
Results, paragraph 2How it will be judged
Counts of a record, so the tolerance is zero; a different count is reported as a difference in the data set, not a failed finding.
- cue-responsive-screen
257 of the 435 neurons (about 59%) were cue-responsive, recorded in 153 sessions, with at least one cue-responsive neuron in each of the 14 rats.
This screen identified 257 cue-responsive neurons (~59% of all recorded neurons) from 153 sessions, with at least one cue-responsive neuron identified in each of the 14 rats
Results, paragraph 2; Methods, Identifying cue-responsive neuronsHow it will be judged
As the Methods give it. For each neuron, cue (danger, uncertainty with shock and omission trials pooled, safety) and window (first 1 s, last 5 s of cue), a paired two-tailed t-test across trials of the window's mean rate (Hz) against the mean rate in the 10 s before that trial's cue onset; a neuron is cue-responsive when any of the six tests has p < 0.05, with no correction (the Methods say Bonferroni was not applied). Reproduced when both counts are within tolerance and every rat has a cue-responsive neuron.
- fig1f-suppression-cue-effect
Suppression ratios were high to danger, intermediate to uncertainty and low to safety, with a main effect of cue in an ANOVA on the rats' mean suppression ratios.
Suppression ratios were high to danger, intermediate to uncertainty, and low to safety (Fig. 1f ). Analysis of variance (ANOVA) for mean individual suppression ratio [factor... revealed a main effect of cue
Results, paragraph 3; Figure 1fHow it will be judged
Suppression ratio per trial as (baseline poke rate - cue poke rate)/(baseline + cue poke rate), baseline the 20 s before cue onset and cue the 10 s cue (Methods, Calculating suppression ratios); averaged per rat and cue over the sessions with cue-responsive neurons, under readings O3. Reproduced when a one-way repeated measures ANOVA on the 14 rats gives a cue effect with p < 0.05 and the rat means order danger > uncertainty > safety, under any reading.
- fig1f-differential-suppression
Suppression ratios differed for every cue pair; the 95% bootstrap intervals of the differences excluded zero for danger vs. uncertainty (0.28), uncertainty vs. safety (0.52) and danger vs. safety (0.80).
The 95% bootstrap confidence interval for differential suppression ratio did not contain zero for danger vs. uncertainty (mean = 0.28, 95% CI [(lower bound) 0.19, (upper bound) 0.38]), uncertainty vs. safety ( M = 0.52, 95% CI [0.35, 0.65]), and danger vs. safety ( M = 0.80, 95% CI [0.65, 0.98]
Results, paragraph 3; Figure 1fHow it will be judged
Per-rat differences of the rat means of fig1f-suppression-cue-effect, bootstrap interval over the 14 rats. Reproduced when, under one reading of O3, all three intervals exclude zero and the three means are within tolerance.
- fig2-excited-inhibited-split
Of the 257 cue-responsive neurons, 131 (about 51%) increased firing to danger (cue-excited) and 126 (about 49%) decreased it (cue-inhibited).
Neurons with positive values (>0) increased danger firing rate over baseline and were designated as cue-excited ( n = 131 ... Neurons with negative values (<0) decreased firing rate below baseline and were designated as cue-inhibited ( n = 126
Results, Diversity in VP baseline firing and threat responding, paragraph 1; Figure 2bHow it will be judged
For each cue-responsive neuron, the danger firing in the first 1 s and in the last 5 s of cue, each relative to the 10 s baseline, averaged into one value; positive is cue-excited, negative cue-inhibited (Results). Open whether the firing is baseline-subtracted Hz or Z-scored firing, and whether the two windows are averaged as two window means or pooled over the 6 s; all four are run. Reproduced when both counts are within tolerance under one reading.
- ancova-excited-no-baseline-effect
For cue-excited neurons, baseline firing rate did not inform the cue firing pattern; an ANCOVA found no baseline x cue x interval interaction.
found no baseline × cue × interval interaction ( F 110,14080 = 1.12, p = 0.18
Results, Diversity in VP baseline firing and threat responding, paragraph 1How it will be judged
A claim of absence, judged on the paper's test. Repeated measures ANCOVA on normalized firing of cue-excited neurons, within factors cue (3) and interval (56 bins of 250 ms, 2 s before cue onset to 2 s after offset), covariate baseline firing rate (Hz, 10 s before cue onset). Reproduced when the baseline x cue x interval interaction has p >= 0.05; not reproduced when p < 0.05. F and p are reported either way, under O1 (a) and (b).
- ancova-inhibited-baseline-effect
For cue-inhibited neurons, baseline firing rate did inform the cue firing pattern; the ANCOVA found a significant baseline x cue x interval interaction.
We separately applied ANCOVA to cue-inhibited neurons to determine if baseline firing rate informed the cue firing pattern. Now, ANCOVA revealed a significant baseline × cue × interval interaction
Results, Diversity in VP baseline firing and threat responding, paragraph 2How it will be judged
The ANCOVA of ancova-excited-no-baseline-effect on the cue-inhibited neurons. Reproduced when the three-way interaction has p < 0.05 under one reading of O1.
- kmeans-three-clusters
K-means clustering of cue-inhibited neurons on baseline firing rate, coefficient of variance, coefficient of skewness, waveform half duration and amplitude ratio gave three clusters differing mainly in baseline firing rate, Low (74), Intermediate (34) and High (18) firing; four of the five features differed across clusters, baseline firing rate most, and amplitude ratio did not.
As a result, cue-inhibited neurons could be divided into three clusters that differed primarily in baseline firing rate: Low firing ( n = 74), Intermediate firing ( n = 34), and High firing ( n = 18) neurons
Results, Diversity in VP baseline firing and threat responding, paragraph 2; Methods, Firing and waveform characteristics, K-means clusteringHow it will be judged
Features as the Methods define them; baseline rate over the 10 s before cue onset; CV and skewness (3 x (mean - median)/SD) of inter-spike intervals; the Methods do not say over which spikes, so both readings are run, (a) the whole session and (b) the 10 s baseline periods; half duration and amplitude ratio from the mean waveform. K-means with k = 3 under readings O7, clusters named by mean baseline rate. Reproduced when, under one reading, the three counts are within tolerance, a one-way ANOVA across clusters has p < 0.05 for baseline rate, CV, skewness and half duration, baseline rate has the largest F, and amplitude ratio has p >= 0.05. Partially reproduced when the counts hold but the feature pattern does not.
- clusters-differ-in-cue-firing
The three cue-inhibited clusters differed in patterned cue firing; each pair gave a significant cluster x cue x interval interaction.
Between-cluster differences in patterned cue firing were confirmed by ANOVA returning a significant cluster × cue × interval interaction for all comparisons
Results, Diversity in VP baseline firing and threat responding, paragraph 2How it will be judged
For each pair of clusters, mixed ANOVA on normalized firing with cluster between and cue and interval (56 bins) within. Reproduced when the three-way interaction has p < 0.005 for all three pairs, under one reading of O1 and O7; partially reproduced when p < 0.05 for all three.
- clusters-by-rat
Low firing neurons came from 11 of 14 rats and Intermediate firing neurons from 9 of 14, while High firing neurons came from only 5 rats, 11 of the 18 from one rat (PA02).
Low firing neurons were observed in 11 of 14 rats and Intermediate firing neurons in 9 of 14 rats ... High firing neurons were observed in only 5 of 14 rats, with 11 of 18 High firing neurons coming from a single rat (PA02
Results, Diversity in VP baseline firing and threat responding, paragraph 3How it will be judged
Counted from the clusters of kmeans-three-clusters and the rat identifiers in the unit names. A tolerance of one rat is used for the rat counts, as 10% of 5 to 11 rats is under one rat.
- fig3a-low-differential-cue-firing
Low firing neurons showed the greatest inhibition to danger, modest inhibition to uncertainty and none to safety, throughout the cue; the ANOVA gave main effects of cue and interval and a cue x interval interaction.
Low firing neurons showed greatest inhibition of firing to danger, modest inhibition to uncertainty and no inhibition to safety. The relative firing pattern was maintained throughout cue presentation
Results, Differential inhibition of firing is maximal to danger, paragraph 1; Figure 3aHow it will be judged
Repeated measures ANOVA on normalized firing, cue (3) x interval (56 bins). Reproduced when cue, interval and cue x interval all have p < 0.05 and the mean normalized firing over the 10 s cue orders danger < uncertainty < safety, under one reading of O1 and O7 (see O6 for n).
- fig3b-low-inhibited-to-threat-cues
Low firing neurons were inhibited below baseline to danger and uncertainty at cue onset and in the late cue period, but not to safety.
Low firing neurons showed selective firing inhibition to threat cues at onset (danger: M = -0.37, 95% CI [-0.44, -0.29]; uncertainty: M = -0.14, 95% CI [-0.22, -0.06]) and during late cue (danger: M = -0.27, 95% CI [-0.34, -0.19]; uncertainty: M = -0.09, 95% CI [-0.15, -0.04]
Results, Differential inhibition of firing is maximal to danger, paragraph 3; Figure 3bHow it will be judged
Per neuron, mean normalized firing in each period and cue; bootstrap intervals over the 74 neurons. "Selective" is read from the paper's own test, the safety intervals at onset and late cue containing zero. Reproduced when the four listed intervals exclude zero below it, the four means are within tolerance, and both safety intervals contain zero, under one reading of O1 and O7.
- fig3b-low-differential
Low firing neurons fired differentially to danger and uncertainty at onset and late cue; uncertainty and safety firing were similar at onset but differed during late cue and delay.
Low firing neurons showed differential firing to danger and uncertainty at onset ( M = -0.23, 95% CI [-0.31, -0.13] ... and during late cue ( M = -0.17, 95% CI [-0.25, -0.09] ... Similar firing was observed to uncertainty and safety at onset ( M = -0.08, 95% CI [-0.19, 0.02] ... but differential firing was observed during late cue ( M = -0.14, 95% CI [-0.25, -0.04] ... and delay ( M = -0.12, 95% CI [-0.26, -6.32 × 10 −4 ]
Results, Differential inhibition of firing is maximal to danger, paragraph 3; Figure 3bHow it will be judged
Per-neuron differences of period means, bootstrap over the 74 neurons. Reproduced when the danger-uncertainty intervals at onset and late cue and the uncertainty-safety intervals at late cue and delay exclude zero, the uncertainty-safety interval at onset contains zero, and the means are within tolerance, under one reading of O1 and O7. The delay interval's upper bound is near zero in the paper, so its exclusion of zero is reported with the seed used.
- fig3c-intermediate-differential-cue-firing
Intermediate firing neurons showed the greatest inhibition to danger at cue onset, less to uncertainty and least to safety, with inhibition specific to danger and uncertainty for the rest of the cue; the ANOVA gave main effects of cue and interval and a cue x interval interaction.
At cue onset, there was greatest firing inhibition to danger, lesser inhibition to uncertainty and least inhibition to safety. Firing inhibition that was specific to danger and uncertainty was maintained for the remainder of cue presentation
Results, Differential inhibition of firing is maximal to danger, paragraph 4; Figure 3cHow it will be judged
As fig3a-low-differential-cue-firing, on the Intermediate firing neurons, with the onset means ordered danger < uncertainty < safety.
- fig3d-intermediate-inhibition
Intermediate firing neurons were inhibited to all three cues at onset but only to danger and uncertainty during late cue.
Intermediate firing neurons were inhibited to all cues at onset (danger: M = -0.54, 95% CI [-0.69, -0.39]; uncertainty: M = -0.25, 95% CI [-0.40, -0.08]; safety: M = -0.21, 95% CI [-0.30, -0.11]), but were selectively inhibited to danger and uncertainty during late cue (danger: M = -0.49, 95% CI [-0.60, -0.36]; uncertainty: M = -0.23, 95% CI [-0.34, -0.12]
Results, Differential inhibition of firing is maximal to danger, paragraph 4; Figure 3dHow it will be judged
As fig3b-low-inhibited-to-threat-cues, on the 34 Intermediate firing neurons. Reproduced when the five listed intervals exclude zero below it, the late-cue safety interval contains zero, and the means are within tolerance, under one reading of O1 and O7.
- fig3d-intermediate-differential
Intermediate firing neurons fired differentially to danger and uncertainty at onset and late cue; uncertainty and safety did not differ at onset but differed during late cue.
Differential inhibition of firing to danger and uncertainty was observed at cue onset ( M = -0.29, 95% CI [-0.44, -0.14] ... and during late cue ( M = -0.26, 95% CI [-0.42, -0.10] ... differential inhibition of firing was not observed to uncertainty and safety at cue onset ( M = -0.05, 95% CI [-0.23, 0.14] ... but was observed during late cue ( M = -0.21, 95% CI [-0.34, -0.05]
Results, Differential inhibition of firing is maximal to danger, paragraph 4; Figure 3dHow it will be judged
As fig3b-low-differential, on the Intermediate firing neurons. Reproduced when the three intervals reported as excluding zero do so, the onset uncertainty-safety interval contains zero, and the means are within tolerance, under one reading of O1 and O7.
- supp5-low-cue-correlations
In Low and Intermediate firing neurons, firing to danger and uncertainty was commonly positively correlated across neurons, while firing to uncertainty and safety showed zero or negative correlations.
positive firing relationships were commonly observed for threat cues, danger and uncertainty, but zero or even negative firing relationships were observed for uncertainty and safety
Results, Differential inhibition of firing is maximal to danger, paragraph 5; Supplementary Fig. 5 (not read)How it will be judged
Pearson correlation across neurons of normalized firing (Methods, Population and single-unit firing analyses), per cluster (Low, Intermediate) and period (onset, late cue, delay), danger against uncertainty and uncertainty against safety. The supplementary figure that gives the panels was not read and the text read does not define "commonly"; it is judged here as at least two of the three periods, an assumption of this file that the outcome rests on. Reproduced when, in each of the two clusters, danger-uncertainty r is positive with p < 0.05 in at least two periods and uncertainty-safety r is never positive with p < 0.05.
- supp6-low-nosepoke-pauses
Pauses in nose poking during the inter-trial interval, in the absence of cues, did not inhibit Low and Intermediate firing neurons, so their cue inhibition is not simply the cessation of nose poking.
Firing inhibition was not simply due to the cessation of nose poking. Pauses in nose poking in the absence of cues during the inter-trial interval were insufficient to inhibit the activity of Low and Intermediate firing neurons
Results, Differential inhibition of firing is maximal to danger, paragraph 5; Supplementary Fig. 6 (not read); Methods, Population and single-unit firing analysesHow it will be judged
A claim of absence, judged on the paper's tests (Methods). Per cluster, repeated measures ANOVA of normalized firing with interval (16 bins of 250 ms, 2 s before to 2 s after nose-poke cessation) as factor, and bootstrap intervals of the post (first 500 ms after cessation) firing against zero and of post minus pre (500 ms before). What counts as a pause is defined in the unread supplement; the readings are a gap of at least (a) 10 s, the cue length, or (b) 5 s, the late-cue window, after a poke, outside the trial window. Neither length comes from the text read, so the outcome rests on an assumption. Normalization as O4. Reproduced when, in each cluster, the post interval and the post minus pre interval do not exclude zero below it; not reproduced when they do. The interval effect of the ANOVA is reported.
- fig4a-low-rt-over-fo
Across all five uncertainty assignments, Low firing neurons signaled relative threat more strongly than fear output (cue-averaged relative-threat betas further from zero than fear-output betas).
Low firing neurons more strongly signaled relative threat, compared to fear output, across all five assignments
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 2; Figure 4aHow it will be judged
Single-unit regression as the Methods give it. For each neuron, its 16 trials, and each 1 s bin from 2 s before cue onset to 2 s after offset, normalized firing (O2) regressed with Matlab regress on a constant, the trial's cue suppression ratio (fear output, O3 for 0/0 trials), and relative threat (danger 1, safety 0, uncertainty 0, 0.25, 0.50, 0.75 or 1, one model per assignment). Threat tuning value = beta averaged over the ten cue bins. Strength is distance from zero (Methods, Single-unit linear regression). Reproduced when, for every assignment, the mean relative-threat beta over Low neurons is further from zero than the mean fear-output beta, under one reading; the signs are reported. At the 1.00 assignment uncertainty and danger share a value; the fit is run as is.
- fig4a-low-025-better-than-000
For Low firing neurons, relative threat signaling was better captured by the actual shock probability (0.25) than by equating uncertainty with safety (0.00); ANOVA on betas gave an assignment x regressor x interval interaction.
Relative threat signaling was better captured by the actual shock probability (assignment = 0.25), compared to an assignment that equated uncertainty to safety (assignment = 0.00)
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 2; Figure 4aHow it will be judged
Repeated measures ANOVA on betas of fig4a-low-rt-over-fo, assignment (0.00, 0.25) x regressor (fear output, relative threat) x interval (14 bins of 1 s). Reproduced when the three-way interaction has p < 0.05 and the mean cue relative-threat beta is further from zero at 0.25 than at 0.00, under one reading of O2 and O3.
- fig4a-low-025-similar-050
For Low firing neurons, relative threat signaling of the actual probability (0.25) was similar to that of the midpoint probability (0.50); ANOVA found no assignment x regressor x interval interaction.
By contrast, ANOVA for 0.25 and 0.50 beta coefficients found no assignment × regressor × interval interaction ( F 13,936 = 0.88, p = 0.57
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 2; Figure 4aHow it will be judged
A claim of absence, judged on the paper's test, the ANOVA of fig4a-low-025-better-than-000 with assignments 0.25 and 0.50. Reproduced when the three-way interaction has p >= 0.05; not reproduced when p < 0.05; F and p reported under every reading of O2 and O3.
- fig4b-low-dynamic-shift
Low firing neurons first decreased firing according to the actual shock probability (0.25) and later according to the midpoint probability (0.50); ANOVA on betas of the two relative-threat regressors gave an assignment x interval interaction.
Low firing neurons initially decreased firing according to the actual shock probability (assignment = 0.25), but later decreased firing according to the greater-than-actual, midpoint probability (assignment = 0.50) (Fig. 4b ). In support, ANOVA found an assignment × interval interaction
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 3; Figure 4bHow it will be judged
Regression "using two relative threat regressors with assignments of 0.25 and 0.50" (Methods, Single-unit linear regression). Open whether the model is (a) constant plus both relative-threat regressors, (b) constant, both relative-threat regressors and fear output, or (c) the 0.25 and 0.50 betas taken from the two separate single-assignment models of fig4a-low-rt-over-fo. Each is run, with O2. Repeated measures ANOVA on betas, assignment (2) x interval (14 bins of 1 s). Reproduced when the interaction has p < 0.05, under one reading.
- fig4c-low-early-late-betas
In Low firing neurons, early betas (first 4 s of cue) were below zero for the 0.25 assignment but not for 0.50, and late betas (last 2 s of cue plus 2 s delay) were below zero for 0.50 but not for 0.25.
early beta coefficients (first 4 s of cue) were shifted below zero for the 0.25 assignment ( M = -0.57, 95% CI [-1.02, -0.10]), but not for the 0.50 assignment ( M = 0.06, 95% CI [-0.38, 0.48] ... late beta coefficients (last 2 s cue plus 2 s delay) were shifted below zero for the 0.50 assignment ( M = -0.51, 95% CI [-1.03, -0.03]), but not for the 0.25 assignment ( M = 0.20, 95% CI [-0.22, 0.62]
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 3; Figure 4cHow it will be judged
Per neuron, the betas of fig4b-low-dynamic-shift averaged over the 1 s bins 0 to 4 s after cue onset (early) and 8 to 12 s (late); bootstrap over neurons. Reproduced when the two intervals reported as excluding zero do so below zero, the other two contain zero, and the means are within tolerance, under one reading.
- fig4d-low-early-late-shift
In Low firing neurons the early-to-late change in beta was positive for the 0.25 assignment and negative for 0.50, and the two shifts differed.
there was a positive early-to-late shift in beta coefficients for the 0.25 assignment ( M = 0.76, 95% CI [0.41, 1.09]), but a negative shift for the 0.50 assignment ( M = -0.56, 95% CI [-0.95, -0.16]), and these shifts differed from one another ( M = 1.32, 95% CI [0.57, 2.04]
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 3; Figure 4dHow it will be judged
Per-neuron late minus early of fig4c-low-early-late-betas, and its difference between assignments; bootstrap over neurons. Reproduced when all three intervals exclude zero with the reported signs and the means are within tolerance, under one reading.
- fig4e-int-025-better-than-000
Intermediate firing neurons signaled relative threat more strongly than fear output, and the actual probability (0.25) did better than 0.00; ANOVA on betas for 0.00 and 0.25 gave an assignment x regressor x interval interaction.
Intermediate firing neurons also more strongly signaled relative threat compared to fear output (Fig. 4e ). Relative threat signaling of the actual probability was superior to an assignment equating uncertainty to safety (assignment = 0.00)
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 4; Figure 4eHow it will be judged
As fig4a-low-rt-over-fo and fig4a-low-025-better-than-000, on the Intermediate firing neurons. Reproduced when the mean cue relative-threat beta is further from zero than the fear-output beta at every assignment, the three-way interaction has p < 0.05, and the 0.25 relative-threat beta is further from zero than the 0.00 one, under one reading.
- fig4e-int-025-similar-050
For Intermediate firing neurons, ANOVA for 0.25 and 0.50 betas found no assignment x regressor x interval interaction.
However, ANOVA for 0.25 and 0.50 beta coefficients revealed no assignment × regressor × interval interaction ( F 13,416 = 1.09, p = 0.37
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 4; Figure 4eHow it will be judged
A claim of absence, judged on the paper's test, as fig4a-low-025-similar-050 on the Intermediate firing neurons. Reproduced when p >= 0.05; not reproduced when p < 0.05.
- fig4f-int-no-assignment-interval
In Intermediate firing neurons, the ANOVA on the 0.25 and 0.50 relative-threat betas found no assignment x interval interaction.
Early signaling of the actual probability (assignment = 0.25) that switched to signaling of the midpoint probability (assignment = 0.50) was only partially observed in Intermediate firing neurons. ANOVA revealed no assignment × interval interaction
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 5; Figure 4fHow it will be judged
A claim of absence, judged on the paper's test, the ANOVA of fig4b-low-dynamic-shift on the Intermediate firing neurons. Reproduced when the interaction has p >= 0.05; not reproduced when p < 0.05; F and p under every reading.
- fig4g-int-early-late-betas
In Intermediate firing neurons early betas were below zero for the 0.25 assignment but not for 0.50, and late betas did not differ from zero for either.
Early beta coefficients were shifted below zero for the 0.25 assignment ( M = -0.77, 95% CI [-1.50, -0.28]), but not for assignment of 0.50 ( M = 0.07, 95% CI [-0.58, 1.14] ... Late beta coefficients were not shifted from zero for either assignment (0.25: M = 0.13, 95% CI [-0.53, 0.76]; 0.50: M = -0.43, 95% CI [-1.05, 0.17]
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 5; Figure 4gHow it will be judged
As fig4c-low-early-late-betas, on the Intermediate firing neurons. Reproduced when only the early 0.25 interval excludes zero (below it) and the means are within tolerance, under one reading.
- fig4h-int-early-late-shift
In Intermediate firing neurons the early-to-late shift was positive for 0.25, not different from zero for 0.50, and the two shifts differed.
There was a positive, early-to-late shift in beta coefficients for the 0.25 assignment ( M = 0.88, 95% CI [0.32, 1.37]), no shift for assignment of 0.50 ( M = -0.52, 95% CI [-1.12, 0.07]), yet these shifts differed from one another ( M = 1.38, 95% CI [0.23, 2.39]
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 5; Figure 4hHow it will be judged
As fig4d-low-early-late-shift, on the Intermediate firing neurons. Reproduced when the 0.25 shift and the difference of shifts exclude zero above it, the 0.50 shift interval contains zero, and the means are within tolerance, under one reading.
- fig4-low-int-combined
Pooling Low and Intermediate firing neurons, ANOVA found an assignment x interval interaction and no neuron type x assignment x interval interaction, so the Intermediate shift did not differ from the Low one.
ANOVA for Low and Intermediate firing neurons combined ... found an assignment × interval interaction ( F 13,1378 = 2.52, p = 0.002 ... but no neuron type × assignment × interval interaction ( F 13,1378 = 0.72, p = 0.74
Results, Low and Intermediate firing neurons dynamically signal relative threat, paragraph 5How it will be judged
Mixed ANOVA on the 0.25 and 0.50 betas of fig4b-low-dynamic-shift, neuron type between, assignment and interval (14 bins) within. Reproduced when assignment x interval has p < 0.05 and the three-way interaction p >= 0.05 (the absence judged on this test alone), under one reading; the three-way F and p reported either way.
- fig5a-low-reward-firing
Low firing neurons sharply increased firing after reward delivery and Intermediate firing neurons did not; ANOVA gave a cluster x interval interaction and a main effect of interval in Low but not Intermediate neurons.
Low firing neurons sharply increased responding following reward delivery, and this firing increase was absent in Intermediate firing neurons
Results, Low firing neurons show opposing responses to threat and reward, paragraph 2; Figure 5aHow it will be judged
Reward-locked normalized firing under readings O4, 16 bins of 250 ms around pellet dispenser advance. Mixed ANOVA cluster (Low, Intermediate) x interval, then one-way repeated measures ANOVA per cluster. Reproduced when cluster x interval has p < 0.05, the Low interval effect has p < 0.05 and the Intermediate one p >= 0.05 (the absence judged on that test), under one reading.
- fig5bc-reward-pre-post
In Low firing neurons pre-reward firing hovered around zero, while post-reward firing exceeded pre-reward firing and zero; in Intermediate firing neurons neither pre- nor post-reward firing differed from zero.
Pre-reward responding by Low firing neurons hovered around zero ( M = -0.06, 95% CI [-0.15, 0.01]), while post-reward firing exceeded pre-reward firing ( M = 0.34, 95% CI [0.15, 0.53]) and differed from zero ( M = 0.28, 95% CI [0.14, 0.42] ... pre-reward and post-reward firing never differed from zero for Intermediate firing neurons (pre-reward: M = 2.60 × 10 −4 , 95% CI [-0.17, 0.15]; post-reward: M = -0.14, 95% CI [-0.41, 0.12]
Results, Low firing neurons show opposing responses to threat and reward, paragraph 2; Figure 5b and 5cHow it will be judged
Pre = the 500 ms before, post = the first 500 ms after dispenser advance (Methods), per neuron, bootstrap over neurons, readings O4. Reproduced when the Low post and post minus pre intervals exclude zero above it, the Low pre and both Intermediate intervals contain zero, and the means are within tolerance, under one reading.
- fig5d-f-low-reward-cue-correlations
Across Low firing neurons, larger reward firing increases went with larger firing decreases to danger and to uncertainty (negative correlations), with no relationship between reward and safety firing.
Low firing neurons showing greater reward firing increases, showed greater danger firing decreases ( R 2 = 0.08, p = 0.01; Fig. 5d ). Reward and uncertainty firing were also negatively correlated ( R 2 = 0.16, p = 4.63 × 10 −4 ; Fig. 5e ), but zero firing relationship was observed for reward and safety ( R 2 = 0.006, p = 0.51
Results, Low firing neurons show opposing responses to threat and reward, paragraph 3; Figure 5d to 5fHow it will be judged
Per neuron, reward value and cue value (second 250 ms bin of cue, trough minus pre) under readings O4 and O5; Pearson correlation over the 74 neurons. Reproduced when, under one reading, reward-danger and reward-uncertainty r are negative with p < 0.05 and reward-safety has p >= 0.05 (absence on the paper's test); partially reproduced when that holds but an R2 is outside its tolerance.
- fig5-low-fisher-comparisons
In Low firing neurons the danger-reward and uncertainty-reward correlations did not differ, while the uncertainty-reward and safety-reward correlations did.
equivalent danger-reward and uncertainty-reward correlations were observed in Low firing neurons (Fisher r-to-z transformation, Z = 0.74, p = 0.46), but uncertainty-reward and safety-reward correlations significantly differed (Fisher r-to-z transformation, Z = 2.96, p = 0.0031
Results, Low firing neurons show opposing responses to threat and reward, paragraph 3How it will be judged
Fisher r-to-z comparison of the correlations of fig5d-f-low-reward-cue-correlations. The paper does not give the formula. The reported values fit the independent-samples form, Z = (z1 - z2)/sqrt(2/(n - 3)) with n = 74 and a positive reward-safety r (from R2 0.08, 0.16 and 0.006 this gives about 0.79 and 2.98), so that is the main reading; a test for dependent correlations sharing the reward variable (Steiger 1980) is shown as a variant. Reproduced when the danger-uncertainty comparison has p >= 0.05 (absence on the paper's test) and the uncertainty-safety comparison has p < 0.05.
- fig5g-i-intermediate-no-reward-cue-correlation
Intermediate firing neurons showed no cue-reward firing relationships, and their correlations did not differ from one another.
No cue-reward firing relationships were observed for Intermediate firing neurons (Fig. 5g-i ), and these correlations did not differ from one another (all Z < 1, all p > 0.3
Results, Low firing neurons show opposing responses to threat and reward, paragraph 3; Figure 5g to 5iHow it will be judged
A claim of absence, judged on the paper's tests, as fig5d-f-low-reward-cue-correlations and fig5-low-fisher-comparisons on the 34 Intermediate firing neurons. Reproduced when all three Pearson correlations have p >= 0.05 and all three Fisher comparisons p >= 0.05; not reproduced when any is p < 0.05. The r values are reported.
- fig6a-excited-differential-cue-firing
Cue-excited neurons increased firing sharply at cue onset, most to danger, less to uncertainty and least to safety, with differential firing through the cue and delay; the ANOVA gave main effects of cue and interval and a cue x interval interaction.
Cue-excited neurons sharply increased activity at onset, with greatest firing to danger, lesser to uncertainty and least to safety. Differential firing continued during the remainder of the cue and through the 2 s delay period
Results, Differential increases in firing are maximal to danger, paragraph 1; Figure 6aHow it will be judged
As fig3a-low-differential-cue-firing, on the cue-excited neurons, with the onset means ordered danger > uncertainty > safety.
- fig6b-excited-period-firing
Cue-excited neurons increased firing to all cues at onset but only to danger and uncertainty in late cue and delay, and fired differentially to danger vs. uncertainty and uncertainty vs. safety in every period.
Firing increases were observed to all cues at onset, but only to the threat cues, danger and uncertainty, in the remaining periods (Fig. 6b ). Furthermore, differential firing was observed to every cue pair in every period: danger vs. uncertainty (onset: M = 0.26, 95% CI [0.19, 0.34], late cue: M = 0.24, 95% CI [0.18, 0.30], and delay: M = 0.19, 95% CI [0.13, 0.26]), and uncertainty vs. safety (onset: M = 0.21, 95% CI [0.12, 0.29], late cue: M = 0.15, 95% CI [0.08, 0.22], and delay: M = 0.22, 95% CI [0.14, 0.31]
Results, Differential increases in firing are maximal to danger, paragraph 2; Figure 6bHow it will be judged
As fig3b-low-inhibited-to-threat-cues and fig3b-low-differential, on the 131 cue-excited neurons. Reproduced when the onset intervals of all three cues and the late-cue and delay intervals of danger and uncertainty exclude zero above it, the late-cue and delay safety intervals contain zero, all six difference intervals exclude zero above it, and the six means are within tolerance, under one reading of O1.
- supp8-excited-cue-correlations
In cue-excited neurons danger and uncertainty firing were positively correlated in all periods, while uncertainty and safety firing were positively correlated at onset, uncorrelated in late cue and negatively correlated in the delay.
Danger and uncertainty firing were positively correlated for all periods ... By contrast, positively correlated firing to uncertainty and safety at cue onset gave way to zero correlation during late cue and negatively correlated during the delay period
Results, Differential increases in firing are maximal to danger, paragraph 2; Supplementary Fig. 8 (not read)How it will be judged
Pearson correlation over cue-excited neurons of period normalized firing (Methods, Population and single-unit firing analyses). Reproduced when danger-uncertainty r is positive with p < 0.05 in all three periods, and uncertainty-safety r is positive with p < 0.05 at onset, p >= 0.05 in late cue, and negative with p < 0.05 in the delay, under one reading of O1.
- supp9-excited-nosepoke-pauses
Cessation of nose poking in the absence of cues did not increase firing of cue-excited neurons.
Cessation of nose poking in the absence of cues was insufficient to increase firing
Results, Differential increases in firing are maximal to danger, paragraph 2; Supplementary Fig. 9 (not read)How it will be judged
A claim of absence, as supp6-low-nosepoke-pauses on the cue-excited neurons. Reproduced when the post and post minus pre intervals do not exclude zero above it; not reproduced when they do.
- fig6-excited-assignment-050
In cue-excited neurons, single-unit activity reflected a mixture of relative threat and fear output, and relative threat was better captured by the midpoint assignment (0.50) than by 0.25 or 0.75, with assignment x regressor interactions in both comparisons.
Relative threat signaling was better captured by the midpoint shock probability (assignment = 0.50), compared to the actual probability (assignment = 0.25) or a higher probability (assignment = 0.75)
Results, Cue-excited neurons signal relative threat and fear output, paragraph 1How it will be judged
Single-assignment regressions as fig4a-low-rt-over-fo on the cue-excited neurons; repeated measures ANOVA on betas, assignment (2) x regressor (2) x interval (12 bins of 1 s from cue onset to 2 s after offset), for 0.25 against 0.50 and 0.50 against 0.75. Reproduced when the 0.25-0.50 assignment x regressor interaction has p < 0.05, the 0.50-0.75 one has p <= 0.05 (the paper prints p = 0.05), and the mean relative-threat beta over these bins is largest at 0.50 among the three, under one reading of O2 and O3.
- fig6c-excited-regressor-interval
In cue-excited neurons, early signaling of elevated relative threat gave way to joint signaling of relative threat and fear output until shock; ANOVA on the 0.50 and fear output betas gave a main effect of regressor and a regressor x interval interaction.
Early signaling of elevated relative threat gave way to joint signaling of relative threat and fear output until shock delivery
Results, Cue-excited neurons signal relative threat and fear output, paragraph 2; Figure 6cHow it will be judged
The 0.50 relative-threat and fear-output betas of the 0.50 single-assignment model, 14 bins of 1 s from 2 s before cue onset to 2 s after offset; repeated measures ANOVA. Reproduced when the regressor main effect and the interaction both have p < 0.05, under one reading of O2 and O3.
- fig6d-excited-betas
In cue-excited neurons, relative-threat (0.50) betas exceeded zero at onset, late cue and delay, fear-output betas exceeded zero in late cue and delay, and the two differed at onset but not in late cue or delay.
Beta coefficients exceeding zero were observed for relative threat (assignment = 0.50) at cue onset ( M = 0.59, 95% CI [0.41, 0.76]), as well as relative threat and fear output during late cue (0.50: M = 0.38, 95% CI [0.21, 0.54]; FO: M = 0.19, 95% CI [0.08, 0.30]) and delay (0.50: M = 0.43, 95% CI [0.26, 0.61]; FO: M = 0.17, 95% CI [0.05, 0.29]; Fig. 6d ). Relative threat and fear output beta coefficients differ from one another during onset ( M = 0.48, 95% CI [0.20, 0.75], but not during the late cue ( M = 0.19, 95% CI [-0.03, 0.44]), and delay ( M = 0.27, 95% CI [-0.007, 0.56]
Results, Cue-excited neurons signal relative threat and fear output, paragraph 2; Figure 6dHow it will be judged
Per neuron, betas of fig6c-excited-regressor-interval averaged over the onset (0 to 1 s), late cue (5 to 10 s) and delay (10 to 12 s) bins; bootstrap over neurons. Reproduced when the five beta intervals exclude zero above it, the fear-output onset interval contains zero, the onset difference excludes zero and the late-cue and delay differences contain it, and the means are within tolerance, under one reading of O2 and O3.
- fig6ef-excited-reward-firing
Cue-excited neurons increased firing to reward; ANOVA gave a main effect of interval, and post-reward firing exceeded zero and pre-reward firing.
In support of salience signaling, cue-excited neurons increased firing to reward ... Firing increases following reward delivery exceeded zero ( M = 0.22, 95% CI [0.07, 0.35]), and also exceeded pre-reward firing ( M = 0.26, 95% CI [0.10, 0.41]
Results, Cue-excited neurons increase firing to reward, paragraph 1; Figure 6e and 6fHow it will be judged
As fig5a-low-reward-firing and fig5bc-reward-pre-post on the cue-excited neurons. Reproduced when the interval effect has p < 0.05, both intervals exclude zero above it, and the means are within tolerance, under one reading of O4.
- fig6g-excited-reward-cue-correlations
Across cue-excited neurons, larger firing increases at reward peak went with larger increases at danger, uncertainty and safety peaks.
Single units showing greater firing increases at reward peak, showed greater firing increases at danger ( R 2 = 0.09, p = 6.44 × 10 −4 ), uncertainty ( R 2 = 0.12, p = 5.75 × 10 −5 ), and safety peak ( R 2 = 0.08, p = 8.20 × 10 −4
Results, Cue-excited neurons increase firing to reward, paragraph 1; Figure 6gHow it will be judged
Reward value as O5; cue value the first 250 ms bin of cue (Figure 6g legend) minus pre as read in O5; Pearson correlation over cue-excited neurons. Reproduced when all three r are positive with p < 0.05 under one reading; partially reproduced when that holds but an R2 is outside tolerance.
- fig6-excited-peak-cue-vs-reward
In cue-excited neurons peak danger firing differed from peak reward firing, while peak uncertainty and safety firing did not.
Finally, peak danger firing ( M = 0.40, 95% CI [0.13, 0.65]), but not peak uncertainty ( M = 0.13, 95% CI [-0.07, 0.35]), and safety ( M = -0.08, 95% CI [-0.30, 0.13]) firing, differed from peak reward firing
Results, Cue-excited neurons increase firing to reward, paragraph 1How it will be judged
Per neuron, the cue peak value minus the reward peak value of fig6g-excited-reward-cue-correlations, bootstrap over neurons. Reproduced when only the danger interval excludes zero and the means are within tolerance, under one reading of O4 and O5.
Pipeline
- Paper
- waiting (not checked)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none