Distributed coding of choice, action and engagement across the mouse brain

Distributed coding of choice, action and engagement across the mouse brain Steinmetz Zatka-Haas Carandini Harris

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Distributed coding of choice, action and engagement across the mouse brain 2019 Nature 10.1038/s41586-019-1787-x Steinmetz2019-distributed-coding-of-choice-action-and-enga Steinmetz2019-distributed-coding-of-choice-action-and-enga.txt, the project's staged copy 1bc676d044b598799ec73aa83ca3c51193ac3a8b not-redistributable neuroscience

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them WHERE the neurons encoding vision, choice, action and behavioral engagement sit - IT IS UNKNOWN HOW THE NEURONS MEDIATING THESE PROCESSES ARE DISTRIBUTED ACROSS BRAIN REGIONS Quotation withheld because this record is not cleared for redistribution. which regions carry signals that predict the CHOSEN action rather than merely accompanying movement - NEURONAL SIGNALS THAT CORRELATE WITH ACTION DO NOT NECESSARILY CORRELATE WITH ACTION SELECTION Quotation withheld because this record is not cleared for redistribution. NEURONS IN NEARLY ALL REGIONS RESPONDED NON-SPECIFICALLY WHEN THE MOUSE INITIATED AN ACTION, BY CONTRAST, NEURONS ENCODING VISUAL STIMULI AND UPCOMING CHOICES OCCUPIED RESTRICTED REGIONS - and CHOICE SIGNALS WERE RARE AND EMERGED WITH INDISTINGUISHABLE TIMING ACROSS REGIONS Quotation withheld because this record is not cleared for redistribution. existence Quotation withheld because this record is not cleared for redistribution. brain-wide distribution of neural coding for vision, action, choice and task engagement Quotation withheld because this record is not cleared for redistribution. electrophysiology electrophysiology Quotation withheld because this record is not cleared for redistribution. mechanisms A recurrent loop between forebrain and midbrain choice areas keeps their choice signals simultaneous despite their different encoding schemes Quotation withheld because this record is not cleared for redistribution. mechanisms Engagement raises subcortical baseline firing toward the threshold at which movements are triggered, so the same stimulus is more likely to release an action Quotation withheld because this record is not cleared for redistribution. mechanisms Arousal-linked hyperpolarization and reduced spiking in cortex sharpens the signal-to-noise ratio of sensory representations, explaining why engagement suppresses neocortical activity Quotation withheld because this record is not cleared for redistribution. the criterion the paper sets itself before measuring anything: TO CARRY SUCH CHOICE-RELATED SIGNALS, A BRAIN REGION MUST CONTAIN NEURONS WHOSE FIRING SELECTIVELY PREDICTS THE CHOSEN ACTION BEFORE THE ACTION OCCURS Quotation withheld because this record is not cleared for redistribution. the design that makes the criterion testable: THE SAME VISUAL STIMULUS COULD THEREFORE LEAD TO EITHER DIRECTION OF TURN, OR TO NO ACTION, ALLOWING US TO DISSOCIATE THE NEURAL CORRELATES OF THE VISUAL STIMULI, OF ACTION INITIATION AND OF ACTION SELECTION Quotation withheld because this record is not cleared for redistribution. the two encodings that differ without differing in time: IN MIDBRAIN NEARLY ALL CHOICE-SELECTIVE NEURONS (53/54, 98%) PREFERRED CONTRALATERAL CHOICES, while forebrain neurons COULD PREFER EITHER CHOICE, giving A DISTINCTIVE 'BILATERAL' ENCODING OF BOTH CHOICES IN THE FOREBRAIN, VERSUS A 'UNILATERAL' ENCODING OF CONTRALATERAL CHOICES IN THE MIDBRAIN Quotation withheld because this record is not cleared for redistribution. the proposed reason engagement helps: ENHANCED ACTIVITY IN SUBCORTICAL AREAS BRINGS ACTIVITY IN THESE REGIONS CLOSER TO THE LEVEL AT WHICH ACTIONS ARE INITIATED, PROVIDING A POTENTIAL MECHANISM FOR INCREASED PROBABILITY OF ACTION IN ENGAGED STATES Quotation withheld because this record is not cleared for redistribution. and what the global pre-movement activity cannot be: THESE SIGNALS MAY COMPRISE FORMS OF COROLLARY DISCHARGE, BUT CANNOT REFLECT SENSORY RE-AFFERENCE AS THEY WERE OBSERVED PRIOR TO MOVEMENT ONSET Quotation withheld because this record is not cleared for redistribution. Sample sizes were not set by a power analysis. what this limits: every effect size and null in the survey: THE SAMPLE SIZES (n = 10 MICE; n = 39 RECORDING SESSIONS; n = 29,134 NEURONS) WERE NOT DETERMINED WITH A POWER ANALYSIS Quotation withheld because this record is not cleared for redistribution. Whether visual responses in the newly implicated 'motor' choice areas would exist in untrained animals cannot be determined from these data. what this limits: the same scope as limitations[7] - whether visual responses in the newly implicated motor choice areas would exist in untrained animals cannot be determined here Quotation withheld because this record is not cleared for redistribution. Circuit mechanisms, omitted regions (cerebellum, brainstem) and generality to other choice tasks all remain for future work. what this limits: the mechanistic and anatomical completeness of the survey: FUTURE WORK WILL BE REQUIRED TO DETERMINE THE CIRCUIT MECHANISMS THAT ENFORCE THESE PRINCIPLES Quotation withheld because this record is not cleared for redistribution. The experimenter was not blinded to the stimulus contrast condition during acquisition or analysis. what this limits: the contrast-condition comparisons: THE EXPERIMENTER WAS NOT BLINDED TO CONTRAST CONDITION EITHER DURING DATA ACQUISITION OR DURING ANALYSIS Quotation withheld because this record is not cleared for redistribution. Six of 39 sessions had to be dropped from the choice-probability control analysis for too few usable trials. what this limits: the choice-probability control analysis: SIX SESSIONS HAD TO BE EXCLUDED FOR HAVING FEWER THAN 10 TRIALS INCLUDE-ABLE IN THE CP ANALYSIS Quotation withheld because this record is not cleared for redistribution. THE FRACTION OF CROSS-VALIDATED VARIANCE THE KERNELS EXPLAINED WAS FREQUENTLY SMALL, EVEN FOR NEURONS WHOSE MEAN RATES THEY ACCURATELY PREDICTED (50.2% for one and 13.6% for another) what this limits: reading 'this neuron encodes action' as a statement about that neuron's spiking rather than about its trial-averaged rate Quotation withheld because this record is not cleared for redistribution. a quarter of the neurons could not be placed - WE RECORDED FROM 29,134 NEURONS, OF WHICH 22,458 WERE LOCALIZABLE TO ONE OF 42 BRAIN REGIONS what this limits: any per-region count or map: WE RECORDED FROM 29,134 NEURONS ... OF WHICH 22,458 WERE LOCALIZABLE TO ONE OF 42 BRAIN REGIONS Quotation withheld because this record is not cleared for redistribution. the training may have made the finding - WHETHER VISUAL NEURONS WOULD BE FOUND these ‘motor’ AREAS IN UNTRAINED ANIMALS IS NOT CLEAR FROM THESE DATA what this limits: reading the visual responses in midbrain and striatum as a property of those structures rather than of trained animals Quotation withheld because this record is not cleared for redistribution. the survey is not the brain - future work must determine HOW THEY EXTEND TO AREAS SUCH AS CEREBELLUM AND BRAINSTEM OMITTED FROM THE CURRENT SURVEY, and THE EXTENT TO WHICH SIMILAR PRINCIPLES GOVERN THE NEURAL CORRELATES OF DIFFERENT CHOICE TASKS what this limits: the brain-wide reading of the survey: future work must determine HOW THEY EXTEND TO AREAS SUCH AS CEREBELLUM AND BRAINSTEM OMITTED FROM THE CURRENT SURVEY Quotation withheld because this record is not cleared for redistribution. the timing result is a null - WE COULD NOT DISTINGUISH THE TIMING OF CHOICE-RELATED SIGNALS BETWEEN THESE REGIONS, and the recurrent loop is offered as the PARSIMONIOUS explanation of that null rather than as a measurement what this limits: reading the recurrent loop as observed rather than as the simplest account of an absence Quotation withheld because this record is not cleared for redistribution. JOINT PERI-EVENT CANONICAL CORRELATION (jPECC) ANALYSIS, A NOVEL EXTENSION OF THE “joint Peristimulus Time Histogram” METHOD MODIFIED TO DETECT CORRELATIONS IN A “communication subspace” BETWEEN TWO POPULATIONS Quotation withheld because this record is not cleared for redistribution. a version of CHOICE PROBABILITY ANALYSIS FOR TASKS WITH MANY STIMULUS CONDITIONS, CALLED COMBINED-CONDITIONS CHOICE PROBABILITY, which QUANTIFIES THE PROBABILITY THAT A NEURON'S SPIKE COUNT WILL BE GREATER ON TRIALS WITH ONE CHOICE THAN TRIALS OF ANOTHER, FOR MATCHED STIMULUS CONDITIONS Quotation withheld because this record is not cleared for redistribution. the visual contrast matched between the conditions being compared eliminates: the activity difference reflecting different visual drive - CONTRALATERAL VISUAL STIMULUS CONTRASTS WERE MATCHED so that DIFFERENCES IN ACTIVITY DO NOT REFLECT DIFFERING VISUAL DRIVE · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. the task-context comparison restricted to trials where the animal did nothing, against passive presentation eliminates: movement explaining the context difference - BOTH VISUAL STIMULI AND BEHAVIORAL REPORTS WERE IDENTICAL, and the conditions were MATCHED FOR VISUAL STIMULATION AND (LACK OF) ACTION · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. the engagement index computed from PRE-STIMULUS activity only eliminates: the index being a movement correlate - WE EXAMINED ONLY PRE-STIMULUS ACTIVITY, WHICH COULD NOT BE CONFLATED WITH THE LARGE NON-SPECIFIC RESPONSES RELATED TO MOVEMENTS · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. the engagement effect tested against pupil diameter, video-detectable movement, previous reward and inter-trial interval eliminates: arousal, movement, reward history and timing - the effect COULD NOT BE FULLY EXPLAINED BY VARIABILITY IN PUPIL DIAMETER, IN OVERT MOVEMENTS DETECTABLE BY VIDEO RECORDINGS, IN THE PRESENCE OF A REWARD ON THE PREVIOUS TRIAL, OR IN THE INTER-TRIAL INTERVAL · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. the kernel test's threshold set from shuffled data eliminates: the encoding counts being a product of the test - WE USED PARAMETERS WHICH GAVE FALSE-POSITIVE ERROR RATES OF 0.33% ON SHUFFLED DATA · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. the whole choice result repeated with a second, independent statistic eliminates: the conclusion depending on the kernel model - a version of CHOICE PROBABILITY ANALYSIS FOR TASKS WITH MANY STIMULUS CONDITIONS was developed and EMPLOYING THIS METHOD, WE OBTAINED SIMILAR RESULTS · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. the analysis restricted to activity before the movement eliminates: activity caused by the movement being read as predicting it - AS OUR QUESTION CONCERNS ACTIVITY PREDICTIVE OF UPCOMING MOVEMENTS, WE APPLIED THIS ANALYSIS ONLY TO PRE-MOVEMENT ACTIVITY · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. A CRITERION FOR WHAT COUNTS AS EVIDENCE, STATED BEFORE ANYTHING IS MEASURED: NEURONAL SIGNALS THAT CORRELATE WITH ACTION DO NOT NECESSARILY CORRELATE WITH ACTION SELECTION - TO CARRY SUCH CHOICE-RELATED SIGNALS, A BRAIN REGION MUST CONTAIN NEURONS WHOSE FIRING SELECTIVELY PREDICTS THE CHOSEN ACTION BEFORE THE ACTION OCCURS Quotation withheld because this record is not cleared for redistribution. AND A WHOLE CLASS OF DESIGNS CANNOT MEET IT: GO/NOGO TASKS CANNOT DISTINGUISH NEURONS THAT NONSPECIFICALLY FIRE FOR ANY ACTION, FROM NEURONS SELECTIVE FOR SPECIFIC CHOICES. Not a result being corrected - a design being retired, and every finding that used it inherits the ambiguity Quotation withheld because this record is not cleared for redistribution. THE SAME STIMULUS AND THE SAME behavior, DIFFERENT ACTIVITY: BOTH VISUAL STIMULI AND BEHAVIORAL REPORTS WERE IDENTICAL, and EVEN THOUGH THE TWO CONDITIONS WERE MATCHED FOR VISUAL STIMULATION AND (LACK OF) ACTION, THEY WERE ACCOMPANIED IN MANY AREAS BY DIFFERENT ACTIVITY, BOTH BEFORE AND AFTER STIMULUS PRESENTATION. The response is not a property of the stimulus Quotation withheld because this record is not cleared for redistribution. and the context effect has OPPOSITE SIGNS in different structures: ACTIVITY IN VISp WAS LOWER IN THE TASK THAN IN THE PASSIVE TRIALS, WHEREAS ACTIVITY IN CP SHOWED THE OPPOSITE MODULATION - WHILE NEOCORTEX AND SENSORY THALAMUS SHOWED A NET DECREASE, OTHER REGIONS SHOWED A CONSISTENT INCREASE. A brain-wide average of this is zero Quotation withheld because this record is not cleared for redistribution. THE GLOBAL SIGNAL IS ABOUT MOVING, NOT ABOUT THE TASK: THE WIDELY DISTRIBUTED ACTIVITY SEEN FOLLOWING TRIAL ONSET WAS THEREFORE PRESENT ONLY WHEN ANIMALS MOVED, AND REGARDLESS OF THE PARTICULAR STIMULUS AND PARTICULAR ACTION Quotation withheld because this record is not cleared for redistribution. and the ordering of areas reverses with the side of the stimulus: WHEN THE MOUSE SUCCESSFULLY SELECTED AN IPSILATERAL STIMULUS, MOST AREAS WERE AGAIN ACTIVATED, BUT VIS AND SCs WERE NOW AMONGST THE LAST AREAS TO RESPOND, RATHER THAN THE FIRST Quotation withheld because this record is not cleared for redistribution. WHAT IS GLOBAL AND WHAT IS NOT, MEASURED RATHER THAN ASSERTED: NEURONS ENCODING ACTION WERE SPREAD THROUGHOUT ALL RECORDED REGIONS, and THE GREAT MAJORITY OF NEURONS ENCODING ACTION DID NOT REQUIRE AN ADDITIONAL CHOICE KERNEL - THEY RESPONDED EQUALLY FOR MOVEMENTS IN EITHER DIRECTION Quotation withheld because this record is not cleared for redistribution. and the negative that makes the positive mean something: CHOICE-SELECTIVE NEURONS WERE NOT FOUND IN VISp, although THIS SET OF REGIONS ENCODING CHOICE OVERLAPPED PARTIALLY WITH THE VISUAL PATHWAY Quotation withheld because this record is not cleared for redistribution. THE SAME VARIABLE ENCODED IN TWO INCOMPATIBLE WAYS IN TWO PLACES: IN MIDBRAIN NEARLY ALL CHOICE-SELECTIVE NEURONS (53/54, 98%) PREFERRED CONTRALATERAL CHOICES, whereas in forebrain A SIZEABLE PROPORTION PREFERRED IPSILATERAL CHOICES (19/48, 40%). A `midbrain encodes choice` tag and a `forebrain encodes choice` tag are the same words for opposite arrangements Quotation withheld because this record is not cleared for redistribution. A NULL DOING THE ARGUMENTATIVE WORK: WE COULD NOT DISTINGUISH THE TIMING OF CHOICE-RELATED SIGNALS BETWEEN THESE REGIONS, AN OBSERVATION PARSIMONIOUSLY EXPLAINED BY A RECURRENT LOOP ACROSS THEM - the loop is the simplest account of an absence, and is labeled as such Quotation withheld because this record is not cleared for redistribution. and the null was sought two ways: the decoding time course COULD NOT BE STATISTICALLY DISTINGUISHED BETWEEN CHOICE-SELECTIVE REGIONS, and jPECC REVEALED NO LAG FOR ACTIVITY CORRELATIONS BETWEEN FRONTAL AND MIDBRAIN AREAS while it DID reveal A CONSISTENT TIME-LAG between visual and frontal cortex Quotation withheld because this record is not cleared for redistribution. THE MODEL EXPLAINS LITTLE OF WHAT IT PREDICTS WELL: THE FRACTION OF CROSS-VALIDATED VARIANCE THE KERNELS EXPLAINED WAS FREQUENTLY SMALL, EVEN FOR NEURONS WHOSE MEAN RATES THEY ACCURATELY PREDICTED - 50.2% for one example neuron and 13.6% for another Quotation withheld because this record is not cleared for redistribution. THE ERROR RATE OF THE CLASSIFICATION IS PRINTED: WE USED PARAMETERS WHICH GAVE FALSE-POSITIVE ERROR RATES OF 0.33% ON SHUFFLED DATA - so every count of encoding neurons in this paper has a stated denominator of error Quotation withheld because this record is not cleared for redistribution. AN INDEX DEFINED BY THE DIFFERENCE BETWEEN TWO CONTEXTS, AND CHECKED AGAINST FOUR CONFOUNDS: the engagement index COULD NOT BE FULLY EXPLAINED BY VARIABILITY IN PUPIL DIAMETER, IN OVERT MOVEMENTS DETECTABLE BY VIDEO RECORDINGS, IN THE PRESENCE OF A REWARD ON THE PREVIOUS TRIAL, OR IN THE INTER-TRIAL INTERVAL Quotation withheld because this record is not cleared for redistribution. ENGAGEMENT MOVES CORTEX AND SUBCORTEX IN OPPOSITE DIRECTIONS: CORRELATES OF ENGAGEMENT ARE CHARACTERIZED BY ENHANCED SUBCORTICAL ACTIVITY AND SUPPRESSED NEOCORTICAL ACTIVITY PRIOR TO VISUAL STIMULUS ONSET - and the authors flag it, ENGAGEMENT-RELATED CORTICAL SUPPRESSION MIGHT SEEM SURPRISING, GIVEN THAT VISUAL CORTEX IS REQUIRED FOR PERFORMANCE OF THIS TASK Quotation withheld because this record is not cleared for redistribution. A QUARTER OF THE DATA HAS NO ADDRESS: WE RECORDED FROM 29,134 NEURONS, OF WHICH 22,458 WERE LOCALIZABLE TO ONE OF 42 BRAIN REGIONS - every per-region number in the paper has that denominator behind it Quotation withheld because this record is not cleared for redistribution. and a majority are modulated, in both directions: A SIZEABLE FRACTION OF ALL NEURONS (60.0%, 13,466 NEURONS) HAD SIGNIFICANT MODULATION OF FIRING RATE DURING THE TASK - MOST OF THESE NEURONS (74.3%) CONSISTENTLY INCREASED THEIR ACTIVITY, BUT A SIZEABLE MINORITY (20.2%) CONSISTENTLY DECREASED Quotation withheld because this record is not cleared for redistribution. WHAT THE GLOBAL SIGNAL CANNOT BE, ON A TIMING ARGUMENT: THESE SIGNALS MAY COMPRISE FORMS OF COROLLARY DISCHARGE, BUT CANNOT REFLECT SENSORY RE-AFFERENCE AS THEY WERE OBSERVED PRIOR TO MOVEMENT ONSET Quotation withheld because this record is not cleared for redistribution. and the finding may belong to the trained animal rather than to the structure: WHETHER VISUAL NEURONS WOULD BE FOUND these ‘motor’ AREAS IN UNTRAINED ANIMALS IS NOT CLEAR FROM THESE DATA Quotation withheld because this record is not cleared for redistribution. claim_type: existence a second reading reaching the same value. Recorded: existence — note: a second reading reaching the same value from a different sentence. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. method: fit each neuron with kernels for vision, action and choice, then ask by a nested cross-validated test whether the kernel under test is REQUIRED, with a declared false-positive rate on shuffled data a second reading DIFFERING from the recorded value. Recorded: electrophysiology — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. phenomenon: activity that appears everywhere in the brain before a movement, in regions with no established role in choosing it a second reading DIFFERING from the recorded value. Recorded: brain-wide distribution of neural coding for vision, action, choice and task engagement — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. builds-on → Guo et al (ref. 45), headplate and recording-chamber implantation method Quotation withheld because this record is not cleared for redistribution. builds-on → Treves and Rolls (ref. 54), sparseness measure, adapted here as the Focality Index Quotation withheld because this record is not cleared for redistribution. extends → The "joint Peristimulus Time Histogram" method (refs 35, 36), extended here to jPECC Quotation withheld because this record is not cleared for redistribution. builds-on → The previously described two-alternative unforced choice task (ref. 3) Quotation withheld because this record is not cleared for redistribution. critiques → go/nogo designs as evidence about choice-selective neurons differs in: what the design can separate - GO/NOGO TASKS CANNOT DISTINGUISH NEURONS THAT NONSPECIFICALLY FIRE FOR ANY ACTION, FROM NEURONS SELECTIVE FOR SPECIFIC CHOICES Quotation withheld because this record is not cleared for redistribution. extends → reports of widespread action correlates in multiple species Quotation withheld because this record is not cleared for redistribution. qualifies → the classical set of choice-related areas (frontal cortex, striatum, SNr, deep superior colliculus) differs in: two regions nobody expected - BUT ALSO UNEXPECTED REGIONS INCLUDING THE MIDBRAIN RETICULAR NUCLEUS AND ZONA INCERTA Quotation withheld because this record is not cleared for redistribution. action —distinct-from→ choice action choice strength of the relation: 1 · note: NEURONAL SIGNALS THAT CORRELATE WITH ACTION DO NOT NECESSARILY CORRELATE WITH ACTION SELECTION Quotation withheld because this record is not cleared for redistribution. go/nogo —contrasted-with→ choice go/nogo choice strength of the relation: 1 · note: GO/NOGO TASKS CANNOT DISTINGUISH NEURONS THAT NONSPECIFICALLY FIRE FOR ANY ACTION, FROM NEURONS SELECTIVE FOR SPECIFIC CHOICES Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span argues for multi-alternative over go/no-go tasks; contrasted-with is loose. Quotation withheld because this record is not cleared for redistribution. midbrain —contrasted-with→ forebrain midbrain forebrain differs in: THE AXIS IS WHAT THEY SHARE AND WHAT THEY DO NOT, BOTH NAMED IN ONE CLAUSE. 'Although Quotation withheld.' Identical on timing, different on code. A bare 'contrasted-with' records the difference and drops the fact that the obvious explanation - one region leading the other - is ruled out in the same sentence. · strength of the relation: 1 · note: ALTHOUGH Quotation withheld - a 'BILATERAL' ENCODING OF BOTH CHOICES IN THE FOREBRAIN, VERSUS a 'UNILATERAL' ENCODING OF CONTRALATERAL CHOICES IN THE MIDBRAIN 2026-08-20 the project re-read worst chars/edge queue): SOUND - the passage is built as this contrast and says so twice - 'By contrast' and 'in the forebrain…'. Both endpoints named, type correct. KEEP. Quotation withheld because this record is not cleared for redistribution. passive —contrasted-with→ Miss passive Miss strength of the relation: 1 · note: BOTH VISUAL STIMULI AND BEHAVIORAL REPORTS WERE IDENTICAL, and yet THEY WERE ACCOMPANIED IN MANY AREAS BY DIFFERENT ACTIVITY, BOTH BEFORE AND AFTER STIMULUS PRESENTATION Quotation withheld because this record is not cleared for redistribution. ANOVA ANOVA Quotation withheld because this record is not cleared for redistribution. ANOVA ANOVA Quotation withheld because this record is not cleared for redistribution. nested ANOVA nested ANOVA Quotation withheld because this record is not cleared for redistribution. chi-squared test chi-squared test Quotation withheld because this record is not cleared for redistribution. t-test t-test Quotation withheld because this record is not cleared for redistribution. Bonferroni correction Bonferroni correction Quotation withheld because this record is not cleared for redistribution. cross-validation cross-validation Quotation withheld because this record is not cleared for redistribution. regression regression Quotation withheld because this record is not cleared for redistribution. logistic regression logistic regression Quotation withheld because this record is not cleared for redistribution. generalized linear model generalized linear model Quotation withheld because this record is not cleared for redistribution. bootstrap bootstrap Quotation withheld because this record is not cleared for redistribution. principal component analysis principal component analysis Quotation withheld because this record is not cleared for redistribution. canonical correlation analysis canonical correlation analysis Quotation withheld because this record is not cleared for redistribution. ROC / Mann-Whitney U statistic ROC / Mann-Whitney U statistic Quotation withheld because this record is not cleared for redistribution. permutation test permutation test Quotation withheld because this record is not cleared for redistribution. two-way ANOVA two-way ANOVA Quotation withheld because this record is not cleared for redistribution. Fisher's exact test Fisher's exact test Quotation withheld because this record is not cleared for redistribution. cross-validated cross-validated Quotation withheld because this record is not cleared for redistribution. Neurons whose firing rose non-specifically at the moment of action initiation were present in nearly every one of the 42 recorded brain regions. Quotation withheld because this record is not cleared for redistribution. Neurons coding the visual stimulus and the upcoming choice were not brain-wide but restricted to particular neocortical, basal ganglia and midbrain regions. Quotation withheld because this record is not cleared for redistribution. Midbrain choice neurons were almost all activated before contralateral turns and suppressed before ipsilateral ones, whereas forebrain choice neurons could prefer either direction. Quotation withheld because this record is not cleared for redistribution. Pre-stimulus activity across the brain predicted whether the mouse would respond on a given trial, with subcortical activity raised and neocortical activity lowered during engagement. Quotation withheld because this record is not cleared for redistribution. THE TIME COURSE OF CHOICE DECODING COULD NOT BE STATISTICALLY DISTINGUISHED BETWEEN CHOICE-SELECTIVE REGIONS IN FRONTAL CORTEX, STRIATUM, AND MIDBRAIN Quotation withheld because this record is not cleared for redistribution. Choice-selective neurons were found in frontal cortex, striatum, substantia nigra pars reticulata, higher-order thalamus and deep superior colliculus, and unexpectedly also in the midbrain reticular nucleus and zona incerta. Quotation withheld because this record is not cleared for redistribution. On trials where the mouse missed a contralateral stimulus and did not act, activity appeared in a visual pathway of visual areas, basal ganglia and midbrain but failed to propagate globally. Quotation withheld because this record is not cleared for redistribution. Sixty per cent of all recorded neurons (13,466) showed significant task-related firing-rate modulation. Quotation withheld because this record is not cleared for redistribution. Canonical-correlation analysis of simultaneous populations showed visual cortex leading frontal cortex and midbrain, but no lead-lag relationship at all between frontal cortex and midbrain. Quotation withheld because this record is not cleared for redistribution. The engagement index separated Go from Miss trials consistently across recordings, an effect not fully explained by pupil diameter, overt movement, previous reward or inter-trial interval. Quotation withheld because this record is not cleared for redistribution. More than half of midbrain choice-selective neurons increased firing before one choice and fell below baseline before the other. Quotation withheld because this record is not cleared for redistribution. NEURONS ENCODING ACTION WERE SPREAD THROUGHOUT ALL RECORDED REGIONS, and THE GREAT MAJORITY DID NOT REQUIRE AN ADDITIONAL CHOICE KERNEL - THEY RESPONDED EQUALLY FOR MOVEMENTS IN EITHER DIRECTION Quotation withheld because this record is not cleared for redistribution. CHOICE-SELECTIVE NEURONS WERE NOT FOUND IN VISp, and were found in TWO SUBCORTICAL NUCLEI WHICH UNEXPECTEDLY CONTAINED NEURONS SELECTIVE FOR CHOICE (MRN AND ZI) Quotation withheld because this record is not cleared for redistribution. IN MIDBRAIN NEARLY ALL CHOICE-SELECTIVE NEURONS (53/54, 98%) PREFERRED CONTRALATERAL CHOICES, while in forebrain A SIZEABLE PROPORTION PREFERRED IPSILATERAL CHOICES (19/48, 40%) Quotation withheld because this record is not cleared for redistribution. THE WIDELY DISTRIBUTED ACTIVITY SEEN FOLLOWING TRIAL ONSET WAS PRESENT ONLY WHEN ANIMALS MOVED, AND REGARDLESS OF THE PARTICULAR STIMULUS AND PARTICULAR ACTION Quotation withheld because this record is not cleared for redistribution. EVEN THOUGH THE TWO CONDITIONS WERE MATCHED FOR VISUAL STIMULATION AND (LACK OF) ACTION, THEY WERE ACCOMPANIED IN MANY AREAS BY DIFFERENT ACTIVITY, BOTH BEFORE AND AFTER STIMULUS PRESENTATION Quotation withheld because this record is not cleared for redistribution. WHILE NEOCORTEX AND SENSORY THALAMUS SHOWED A NET DECREASE IN PRE-STIMULUS ACTIVITY DURING TASK CONTEXT, OTHER REGIONS SHOWED A CONSISTENT INCREASE Quotation withheld because this record is not cleared for redistribution. the time course of choice decoding COULD NOT BE STATISTICALLY DISTINGUISHED BETWEEN CHOICE-SELECTIVE REGIONS IN FRONTAL CORTEX, STRIATUM, AND MIDBRAIN Quotation withheld because this record is not cleared for redistribution. WE RECORDED FROM 29,134 NEURONS, OF WHICH 22,458 WERE LOCALIZABLE TO ONE OF 42 BRAIN REGIONS Quotation withheld because this record is not cleared for redistribution. A SIZEABLE FRACTION OF ALL NEURONS (60.0%, 13,466 NEURONS) HAD SIGNIFICANT MODULATION OF FIRING RATE DURING THE TASK - 74.3% CONSISTENTLY INCREASED, 20.2% CONSISTENTLY DECREASED Quotation withheld because this record is not cleared for redistribution. THE FRACTION OF CROSS-VALIDATED VARIANCE THE KERNELS EXPLAINED WAS FREQUENTLY SMALL, EVEN FOR NEURONS WHOSE MEAN RATES THEY ACCURATELY PREDICTED Quotation withheld because this record is not cleared for redistribution. choice choice Quotation withheld because this record is not cleared for redistribution. choice-related signal choice-related signal Quotation withheld because this record is not cleared for redistribution. combined-conditions Choice Probability (ccCP) combined-conditions Choice Probability (ccCP) Quotation withheld because this record is not cleared for redistribution. joint Peri-Event Canonical Correlation (jPECC) joint Peri-Event Canonical Correlation (jPECC) Quotation withheld because this record is not cleared for redistribution. engagement index engagement index Quotation withheld because this record is not cleared for redistribution. Focality index Focality index Quotation withheld because this record is not cleared for redistribution. a statistic that QUANTIFIES THE PROBABILITY THAT A NEURON'S SPIKE COUNT WILL BE GREATER ON TRIALS WITH ONE CHOICE THAN TRIALS OF ANOTHER, FOR MATCHED STIMULUS CONDITIONS, JUST AS IN CLASSIC CHOICE PROBABILITY combined-conditions choice probability Quotation withheld because this record is not cleared for redistribution. ‘Action’ kernel ‘Action’ kernel Quotation withheld because this record is not cleared for redistribution. ‘Choice’ kernel ‘Choice’ kernel Quotation withheld because this record is not cleared for redistribution. contrast condition contrast condition Quotation withheld because this record is not cleared for redistribution. firing rate firing rate Quotation withheld because this record is not cleared for redistribution. contralateral visual stimulus contrast contralateral visual stimulus contrast Quotation withheld because this record is not cleared for redistribution. choice choice Quotation withheld because this record is not cleared for redistribution. engagement engagement Quotation withheld because this record is not cleared for redistribution. pupil area pupil area Quotation withheld because this record is not cleared for redistribution. motion energy motion energy Quotation withheld because this record is not cleared for redistribution. previous trial reward outcome previous trial reward outcome Quotation withheld because this record is not cleared for redistribution. inter-trial interval inter-trial interval Quotation withheld because this record is not cleared for redistribution. 29,134 — neurons recorded in total, of which 22,458 could be assigned to one of 42 brain regions Quotation withheld because this record is not cleared for redistribution. 60.0% — share of recorded neurons with significant task-related firing-rate modulation Quotation withheld because this record is not cleared for redistribution. 53/54, 98% — proportion of midbrain choice-selective neurons preferring contralateral choices Quotation withheld because this record is not cleared for redistribution. 19/48, 40% — proportion of forebrain choice-selective neurons preferring ipsilateral choices, significantly more than in midbrain Quotation withheld because this record is not cleared for redistribution. 0.33% — false-positive rate of the kernel selectivity test estimated on shuffled data Quotation withheld because this record is not cleared for redistribution. 0.079 / 0.069 / 0.040 — THIS FOCALITY INDEX WAS 0.079 FOR CHOICE, 0.069 FOR VISUAL KERNELS AND 0.040 FOR ACTION KERNELS - the differences between Choice and Move, and between Contralateral Vision and Action, WERE STATISTICALLY SIGNIFICANT Quotation withheld because this record is not cleared for redistribution. 1.7 ± 2.5% — incorrect-choice rate on high-contrast single-stimulus trials, with 10.1% misses Quotation withheld because this record is not cleared for redistribution. 94.1% ± 0.6% — proportion of passive-replay trials with high-contrast stimuli in which the mouse made no wheel movement Quotation withheld because this record is not cleared for redistribution. 88.7 μm — median absolute error of the longitudinal probe-to-atlas alignment, close to the 70 micrometer probe shank width Quotation withheld because this record is not cleared for redistribution. 29/54, 54% — midbrain choice-selective neurons showing directionally opposed activity (enhanced for one choice, suppressed below baseline for the other) Quotation withheld because this record is not cleared for redistribution. 53/54, 98% — midbrain choice-selective neurons preferring contralateral choices depends on: the structure - the matched forebrain figure is 19/48, 40% PREFERRING IPSILATERAL CHOICES Quotation withheld because this record is not cleared for redistribution. 0.33% — the FALSE-POSITIVE ERROR RATE ON SHUFFLED DATA that the kernel test's parameters were set to give Quotation withheld because this record is not cleared for redistribution. 29,134 — neurons recorded, OF WHICH 22,458 WERE LOCALIZABLE TO ONE OF 42 BRAIN REGIONS Quotation withheld because this record is not cleared for redistribution. 60.0%, 13,466 neurons — the fraction with SIGNIFICANT MODULATION OF FIRING RATE DURING THE TASK Quotation withheld because this record is not cleared for redistribution. 50.2% ... 13.6% — cross-validated variance explained for two example neurons whose mean rates the model predicted accurately Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.1038/s41586-019-1787-x.