Neural underpinnings of the evidence accumulator

Neural underpinnings of the evidence accumulator Brody Hanks

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.

Neural underpinnings of the evidence accumulator 2016 Current Opinion in Neurobiology 10.1016/j.conb.2016.01.003 Brody2016-neural-underpinnings-of-the-evidence-accumul Brody2016-neural-underpinnings-of-the-evidence-accumul.txt, the project's staged copy 518954f3b6d4dd753d21165e155bcc981ba795b3 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 Which brain regions actually compute the gradual accumulation of evidence, as opposed to merely representing its value. Quotation withheld because this record is not cleared for redistribution. critique Quotation withheld because this record is not cleared for redistribution. gradual accumulation of evidence in perceptual decision-making Quotation withheld because this record is not cleared for redistribution. review review Quotation withheld because this record is not cleared for redistribution. mechanisms When evidence is weak the accumulator drifts only slowly away from the decision boundary, so it takes longer to establish which side it is on, which is why difficult decisions are slow. Quotation withheld because this record is not cleared for redistribution. mechanisms FOF acts downstream of accumulation, categorizing the accumulator's value into a choice, which is why its activity is required near the go cue rather than throughout the stimulus. Quotation withheld because this record is not cleared for redistribution. mechanisms Discrepant PPC inactivation results across rodent visual tasks may arise because the inactivation spreads beyond PPC into neighbouring small secondary visual areas that supply sensory input. Quotation withheld because this record is not cleared for redistribution. Encoding was inferred from ramping firing rates that rise faster on easy trials than hard ones, exactly as the accumulator model predicts Quotation withheld because this record is not cleared for redistribution. Tuning curves separate two regions that look alike in averages: one reads out the graded value, the other a categorical side of the boundary Quotation withheld because this record is not cleared for redistribution. A region can be required at a specific moment rather than throughout, which optogenetic timing can test and lesions cannot Quotation withheld because this record is not cleared for redistribution. Microstimulation cannot localise its own effect, because it acts largely on axons passing through rather than on the cells at the site Quotation withheld because this record is not cleared for redistribution. The proposed test for circuit membership is three conditions at once - inactivation matters, it matters during the accumulation window specifically, and the graded value is encoded there Quotation withheld because this record is not cleared for redistribution. Where a label's boundaries are uncertain, an inactivation aimed at one region may be silencing its neighbour, which would reconcile the conflicting results across species Quotation withheld because this record is not cleared for redistribution. The review restricts itself to two-alternative decisions and excludes human work. what this limits: the scope of the review: WE LIMIT OURSELVES TO TWO-ALTERNATIVE DECISIONS ... AND WILL NOT ADDRESS WORK IN HUMANS Quotation withheld because this record is not cleared for redistribution. Whether rats and primates in fact use the same circuits and algorithms for accumulation is inferred from similar ramping responses and is not established. what this limits: the rat-to-primate inference: WHETHER THIS IS INDEED THE CASE REMAINS AN OPEN QUESTION Quotation withheld because this record is not cleared for redistribution. Because the anatomical border of rodent PPC with adjacent small V2 areas is still uncertain, inactivation effects attributed to PPC may reflect spillover into V2. what this limits: every inactivation result attributed to rodent PPC: THE PRECISE DEFINITION OF RODENT PPC AND THE LOCATION OF ITS BORDER WITH THESE V2 AREAS [is] STILL HIGHLY UNCERTAIN AND A MATTER OF ACTIVE RESEARCH Quotation withheld because this record is not cleared for redistribution. The authors report that their own earlier serial model of the two regions was wrong what this limits: the authors' own earlier account: AS WE DESCRIBE BELOW, THIS SERIAL MODEL TURNED OUT TO BE WRONG Quotation withheld because this record is not cleared for redistribution. Neither of the two candidate regions satisfies the criteria the authors themselves propose what this limits: the two candidate regions against the review's own test: NEITHER THE FOF NOR THE PPC SATISFY ALL THREE CRITERIA Quotation withheld because this record is not cleared for redistribution. A negative inactivation result cannot rule out a redundant circuit in which no single region is required what this limits: the negative inactivation results: the PPC COULD, PERHAPS, PLAY A ROLE ... AS PART OF A MUCH LARGER REDUNDANT CIRCUIT, which a null cannot exclude Quotation withheld because this record is not cleared for redistribution. The region name itself is unsettled, so results labelled PPC may not concern the same tissue across studies what this limits: the comparability of PPC results across studies: FURTHER CLARITY ... WILL GREATLY BENEFIT FROM BETTER DEFINITIONS AND KNOWLEDGE ABOUT THE TOPOGRAPHY OF BRAIN REGIONS NEAR THE LOCATION CURRENTLY REFERRED TO AS RODENT PPC Quotation withheld because this record is not cleared for redistribution. Several of the supporting results are unpublished data from the authors' own laboratory what this limits: several supporting claims: they rest on unpublished data from the authors' own laboratory Quotation withheld because this record is not cleared for redistribution. The causal evidence in primates rests on three microstimulation studies, all with the confound the authors name what this limits: the causal evidence for these areas in primates: ONLY THREE EXISTING PUBLISHED STUDIES, ALL USING ONLY UNILATERAL ELECTRICAL MICROSTIMULATION Quotation withheld because this record is not cleared for redistribution. PPC PPC Quotation withheld because this record is not cleared for redistribution. accumulation of evidence accumulation of evidence Quotation withheld because this record is not cleared for redistribution. a region's involvement - ENCODING: the accumulator's graded value can be read out of its firing, which PPC satisfies Quotation withheld because this record is not cleared for redistribution. a region's involvement - CAUSAL: silencing it changes behaviour, which PPC does not satisfy despite encoding the value Quotation withheld because this record is not cleared for redistribution. necessity - AT ALL: silencing the region changes performance, criterion one Quotation withheld because this record is not cleared for redistribution. necessity - AT THE RIGHT TIME: silencing during the accumulation window specifically changes performance, criterion two, which separates accumulating from reading out Quotation withheld because this record is not cleared for redistribution. PPC - as a name for tissue defined by location, which differs across labs and species Quotation withheld because this record is not cleared for redistribution. PPC - as a name the authors propose reserving for tissue defined by function and connectivity, including the very null result at issue Quotation withheld because this record is not cleared for redistribution. claim_type: critique a second reading reaching the same value. Recorded: critique — 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. phenomenon: gradual accumulation of evidence in perceptual decision-making a second reading reaching the same value. Recorded: gradual accumulation of evidence in perceptual decision-making — 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. builds-on → Brunton et al. (Poisson clicks rat task) Quotation withheld because this record is not cleared for redistribution. critiques → the proposal of a one-to-one relationship between parietal firing rates and the accumulator value Quotation withheld because this record is not cleared for redistribution. critiques → microstimulation as a means of localising causal contributions Quotation withheld because this record is not cleared for redistribution. critiques → the authors' own earlier serial model in which PPC computes the accumulator and FOF reads it out Quotation withheld because this record is not cleared for redistribution. extends → Brunton et al.'s rat Poisson clicks model of accumulation Quotation withheld because this record is not cleared for redistribution. posterior parietal cortex —distinct-from→ evidence accumulation posterior parietal cortex evidence accumulation strength of the relation: 0.9 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the PPC 'MAY play little to no role' in accumulation; distinct-from is hedged. Quotation withheld because this record is not cleared for redistribution. ramping —operationalised-as→ evidence accumulation ramping evidence accumulation strength of the relation: 0.7 Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span proposes a 1-to-1 relation between PPC firing and the accumulator; operationalised-as is loose. Quotation withheld because this record is not cleared for redistribution. frontal orienting fields —causes→ decision boundary frontal orienting fields decision boundary strength of the relation: 0.9 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span 'SUGGESTS' the FOF categorizes the accumulator's value to drive the choice; causes is hedged. Quotation withheld because this record is not cleared for redistribution. eNpHR3.0 —operationalised-as→ optogenetic inactivation eNpHR3.0 optogenetic inactivation strength of the relation: 1 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but eNpHR3.0 is the TOOL used for optogenetic inactivation; operationalised-as is loose. Quotation withheld because this record is not cleared for redistribution. RETRACTED (NEITHER ENDPOINT IS IN THE SPAN, AND THE RELATION IS HEDGED THREE TIMES) · frontal orienting fields —same-as→ frontal eye fields frontal orienting fields frontal eye fields RETRACTED: NEITHER ENDPOINT IS IN THE SPAN, AND THE RELATION IS HEDGED THREE TIMES. The edge is 'frontal orienting fields --same-as--> frontal eye fields'. The span: 'HAD BEEN SUGGESTED…' The span names 'PPC' and 'FEF'; it does not contain either endpoint name as words at all - the endpoints were supplied from the tagger's vocabulary, THE SAME FAULT AS Gelbard-Sagiv2008's absent 'reactivation'. This edge PREDATES the gate check and could not be written today. On top of that: 'suggested as', 'analogous', and 'perhaps even' are three hedges in eleven words, and 'respectively' means the sentence maps TWO pairs at once, so even the intended relation is half of a two-way mapping. — strength of the relation: 0.7 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the FOF was SUGGESTED as analogous or perhaps homologous to the FEF; same-as is hedged. Quotation withheld because this record is not cleared for redistribution. lateral intraparietal —part-of→ posterior parietal cortex lateral intraparietal posterior parietal cortex strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. secondary visual cortex —distinct-from→ posterior parietal cortex secondary visual cortex posterior parietal cortex strength of the relation: 0.9 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but secondary visual cortex lies immediately POSTERIOR TO the PPC; distinct-from is loose. Quotation withheld because this record is not cleared for redistribution. click-triggered average —part-of→ Poisson clicks click-triggered average Poisson clicks strength of the relation: 1 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the click-triggered average is an analysis OF responses to single clicks; part-of is loose. Quotation withheld because this record is not cleared for redistribution. decision boundary —part-of→ evidence accumulation decision boundary evidence accumulation strength of the relation: 1 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but endpoint by anaphora: 'this model' = accumulation of evidence (previous sentence); part-of is loose. Quotation withheld because this record is not cleared for redistribution. PPC —distinct-from→ accumulation of evidence PPC accumulation of evidence the relation itself is separately evidenced · strength of the relation: 1 · note: RE-GROUNDS the record's 'posterior parietal cortex --distinct-from--> evidence accumulation'. The SPAN was always right; the ENDPOINT NAMES did not match its words. 'normalised' maps the abbreviation and the paper's phrasing onto the existing concepts. · the relation stated: the region ENCODES the accumulator's value yet is dispensable for decisions that use it Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the PPC 'MAY play little to no role' in accumulation; distinct-from is hedged. Quotation withheld because this record is not cleared for redistribution. Contrary to expectations, inactivating rat PPC had negligible effects on Poisson Clicks performance. Quotation withheld because this record is not cleared for redistribution. The authors conclude that PPC is the least likely of the regions examined to be a genuine node of the accumulation circuit. Quotation withheld because this record is not cleared for redistribution. The serial account in which PPC computes the accumulator and hands it to FOF was falsified by the perturbation results. Quotation withheld because this record is not cleared for redistribution. Frontal orienting fields carry a categorical, step-like read-out of which side of the decision boundary the accumulator is on, rather than its graded value. Quotation withheld because this record is not cleared for redistribution. Unpublished data suggest at least one subcortical region appears to meet all three criteria and may become the first identified node of the causal accumulation circuit. Quotation withheld because this record is not cleared for redistribution. The authors propose a nomenclature reform, restricting the label 'rodent PPC' to tissue whose physiology and connectivity match primate PPC. Quotation withheld because this record is not cleared for redistribution. silencing parietal cortex had, against expectations, negligible effects on task performance Quotation withheld because this record is not cleared for redistribution. the same inactivations did affect interspersed free-choice trials, so the manipulation itself worked Quotation withheld because this record is not cleared for redistribution. parietal cortex is the only region whose silencing leaves task behaviour unaffected Quotation withheld because this record is not cleared for redistribution. tuning curves were smooth graded functions of accumulator value in PPC but step-like in FOF Quotation withheld because this record is not cleared for redistribution. the mapping from accumulator value to firing rate stays fixed across the trial even as the value changes Quotation withheld because this record is not cleared for redistribution. neither candidate region satisfies all three of the authors' criteria for causal circuit membership Quotation withheld because this record is not cleared for redistribution. evidence accumulation evidence accumulation Quotation withheld because this record is not cleared for redistribution. decision boundary decision boundary Quotation withheld because this record is not cleared for redistribution. tuning curve tuning curve Quotation withheld because this record is not cleared for redistribution. initial criteria initial criteria Quotation withheld because this record is not cleared for redistribution. Poisson clicks task Poisson clicks task Quotation withheld because this record is not cleared for redistribution. secondary visual cortex secondary visual cortex Quotation withheld because this record is not cleared for redistribution. tuning curves tuning curves Quotation withheld because this record is not cleared for redistribution. three — number of published causal perturbation studies of accumulation-related areas in the primate random-dot-motion task at the time of writing Quotation withheld because this record is not cleared for redistribution. 6 — number of rat brain regions whose silencing had been tested on the Poisson Clicks task, of which only PPC showed no behavioural impact Quotation withheld because this record is not cleared for redistribution. 50 — illustrative number of trained subjects needed to deploy a new perturbation technique, used to argue that rodents make perturbation work feasible Quotation withheld because this record is not cleared for redistribution. three initial criteria — the proposed test for whether a region belongs to the causal circuit Quotation withheld because this record is not cleared for redistribution. 6 different rat regions — how many have been silenced, of which only one showed no effect Quotation withheld because this record is not cleared for redistribution. only three existing published studies — the entire causal literature on these primate regions in this task, all by one method 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/10.1016/j.conb.2016.01.003.