Task difficulty and performance induce diverse adaptive patterns in gain and shape of primary auditory cortical receptive fields

Task difficulty and performance induce diverse adaptive patterns in gain and shape of primary auditory cortical receptive fields Atiani Elhilali David Fritz Shamma

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.

Task difficulty and performance induce diverse adaptive patterns in gain and shape of primary auditory cortical receptive fields 2009 Neuron 10.1016/j.neuron.2008.12.027 Atiani2009-task-difficulty-and-performance-induce-diver Atiani2009-task-difficulty-and-performance-induce-diver.txt, the project's staged copy 16ce9d1ed767520ea6cd1618c5d9835e6d5a80d8 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 rapid changes in A1 spectrotemporal receptive fields while an animal detects a tone in noise Quotation withheld because this record is not cleared for redistribution. whether the magnitude of task-driven receptive field plasticity depends on task difficulty, and whether the change is one of gain or of shape Quotation withheld because this record is not cleared for redistribution. separates gain from shape change and shows they move in opposite directions, with the sign of the net effect set jointly by task difficulty and by the cell's distance from the target frequency Quotation withheld because this record is not cleared for redistribution. mechanism Quotation withheld because this record is not cleared for redistribution. rapid task-related plasticity of auditory cortical receptive fields Quotation withheld because this record is not cleared for redistribution. electrophysiology electrophysiology Quotation withheld because this record is not cleared for redistribution. mechanisms Lowering target SNR raises the attentional effort the animal must invest, and that increased effort drives larger adaptive changes in the receptive field Quotation withheld because this record is not cleared for redistribution. mechanisms Pulling a tone out of a masker requires simultaneously suppressing the responses driven by the concurrent noise and enhancing sensitivity to the tone Quotation withheld because this record is not cleared for redistribution. mechanisms Which direction a cell changes is set by what drives it: near cells are enhanced by proximity to the target tone, far cells are suppressed because the behaviourally relevant broadband masker drives them Quotation withheld because this record is not cleared for redistribution. a contrast matched filter: the acoustic features of the foreground are enhanced and those of the background suppressed Quotation withheld because this record is not cleared for redistribution. the net effect keeps the target salient by two different routes - enhancing near cells when the tone is loud, suppressing far cells when it is buried Quotation withheld because this record is not cleared for redistribution. the active STRF is modelled explicitly as a scaled passive STRF plus an orthogonal shape change, S_d = g.S_b + delta Quotation withheld because this record is not cleared for redistribution. STRFs are measured from the reference TORCs alone, never during target presentation, so the measurement does not include the target Quotation withheld because this record is not cleared for redistribution. attentional load is validated behaviourally by inserting identical intermediate-difficulty probes into easy and hard blocks Quotation withheld because this record is not cleared for redistribution. STRF frequency resolution is too coarse to exclude that some A1 neurons build narrow matched filters at the exact target frequency. Quotation withheld because this record is not cleared for redistribution. It is unknown how far the gain-reduction/shape-enhancement pattern is specific to the tone-in-noise task, its avoidance paradigm, or the fixed noise level. Quotation withheld because this record is not cleared for redistribution. Most cells could not be held across multiple task conditions, so difficulty comparisons are largely between rather than within cells. Quotation withheld because this record is not cleared for redistribution. Disentangling stimulus structure from behavioural paradigm requires further experiments; the interpretations offered are conjectures. Quotation withheld because this record is not cleared for redistribution. The 0.6-octave boundary dividing near from far cells was an arbitrary choice. Quotation withheld because this record is not cleared for redistribution. the near/far dividing line is arbitrary, stated as such, and shown not to change the results Quotation withheld because this record is not cleared for redistribution. the standard STRF power normalisation used in the group's previous papers does not separate gain from shape Quotation withheld because this record is not cleared for redistribution. the dependence of gain on task difficulty was NOT confirmed when all performance levels were pooled Quotation withheld because this record is not cleared for redistribution. changes often persisted after the task with no systematic relation to the type of plasticity Quotation withheld because this record is not cleared for redistribution. most cells were recorded in only one task condition, so within-cell comparison across difficulties is limited Quotation withheld because this record is not cleared for redistribution. the finding that harder tasks recruit more attention, and the probe-trial design used to demonstrate it Quotation withheld because this record is not cleared for redistribution. the matched filter, imported as the model for how a receptive field should adapt to separate foreground from background Quotation withheld because this record is not cleared for redistribution. the target tone frequency and signal-to-noise ratio held constant throughout a given task eliminates: a within-task change in the stimulus itself explaining the receptive-field change attributed to task difficulty · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. shape shape Quotation withheld because this record is not cleared for redistribution. plasticity here is rapid and reversible - measured as the difference between active and passive STRFs within a session, not as learning Quotation withheld because this record is not cleared for redistribution. task difficulty is operationalised as target SNR, and validated separately by lick withdrawal time and by probe-trial performance Quotation withheld because this record is not cleared for redistribution. gain and shape are given distinct definitions by an explicit model: gain is the fraction of the STRF that keeps its original shape, shape change is the orthogonal remainder Quotation withheld because this record is not cleared for redistribution. near and far are defined by a 0.6-octave cut from the target frequency, which the paper calls arbitrary and tests Quotation withheld because this record is not cleared for redistribution. method: STRFs measured from responses to modulated broadband TORC references during and before a conditioned-avoidance tone-in-noise detection task, at target SNRs from +15 to -10 dB 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: attention-driven receptive field plasticity in primary auditory cortex a second reading DIFFERING from the recorded value. Recorded: rapid task-related plasticity of auditory cortical receptive fields — 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. extends → Fritz et al. (2003; 2005; 2007a,b), the group's earlier demonstrations of rapid task-related STRF plasticity with pure-tone targets Quotation withheld because this record is not cleared for redistribution. builds-on → the "contrast matched filter" hypothesis of Fritz et al. (2007a) Quotation withheld because this record is not cleared for redistribution. extends → visual studies of task difficulty and attention (Boudreau et al., 2006; Spitzer et al., 1988; Spitzer and Richmond, 1991), which did not separate gain from shape differs in: THE AXIS IS GAIN AGAINST SHAPE. 'Quotation withheld' Two ways a receptive field can change that a single measure cannot separate - so the earlier result is a sum of two effects that this paper's [7] then shows move differently with difficulty. Quotation withheld because this record is not cleared for redistribution. replicates → earlier task-related plasticity results from the same group in tone-versus-TORC and tone-versus-tone paradigms (Fritz et al., 2003; 2005; 2007b) Quotation withheld because this record is not cleared for redistribution. extends → Fritz et al. on rapid STRF plasticity with pure-tone targets Quotation withheld because this record is not cleared for redistribution. qualifies → the contrast matched filter model, by showing the sign of the change depends on difficulty and on distance from target differs in: THE AXIS IS DISTANCE FROM THE TARGET FREQUENCY, AND THE SIGN FLIPS ACROSS IT. 'when we grouped the plasticity changes by near and far cells, it became apparent that near cells on average were enhanced (red) whereas far cells were suppressed (blue).' The matched-filter model is qualified rather than refuted: it holds, with a sign that depends on where the cell sits. Quotation withheld because this record is not cleared for redistribution. extends → visual studies showing task difficulty enhances attentional modulation, which did not separate gain from shape differs in: THE AXIS IS GAIN AGAINST SHAPE, on a second target. Same sentence as [2]: 'THESE MEASUREMENTS DID NOT DISTINGUISH RECEPTIVE FIELD GAIN FROM SHAPE CHANGES.' One limitation of one earlier method, disqualifying it for two different claims this paper makes - which is why the sentence supports two rows rather than one. Quotation withheld because this record is not cleared for redistribution. qualifies → their own gain-versus-difficulty result, which did not survive pooling across performance levels differs in: THE AXIS IS WHETHER THE CELLS ARE POOLED ACROSS PERFORMANCE LEVELS. 'Quotation withheld.' The paper's own effect survives one analysis choice and not another, and it says which - an analyst decision reported as a result rather than buried in methods. Quotation withheld because this record is not cleared for redistribution. EVIDENCE CORRECTED (RE-GROUNDED) · task difficulty —operationalised-as→ signal-to-noise ratio task difficulty signal-to-noise ratio EVIDENCE CORRECTED: RE-GROUNDED. The old span named neither endpoint; this clause states the operationalisation outright - difficulty IS the signal-to-noise ratio in this design. — strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. gain —contrasted-with→ shape gain shape strength of the relation: 1 · note: the paper's central result: they move in OPPOSITE directions in the same cells - gain decreased while shape changes were excitatory Quotation withheld because this record is not cleared for redistribution. task difficulty —causes→ attention task difficulty attention strength of the relation: 0.9 · note: and it is tested rather than assumed, via probe trials of identical difficulty in easy and hard blocks Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span states a hypothesis about attentional resources; causes states it as a finding. Quotation withheld because this record is not cleared for redistribution. attention —operationalised-as→ probe trials attention probe trials strength of the relation: 0.9 Audit 2026-09-28: DOUBTFUL - read against its span and kept, but probe trials were inserted to test attention; operationalised-as is loose. Quotation withheld because this record is not cleared for redistribution. task difficulty —operationalised-as→ signal-to-noise ratio task difficulty signal-to-noise ratio strength of the relation: 1 · note: RE-GROUNDS another row of this record: the relation was always stated, the endpoint names were not in the span. Quotation withheld because this record is not cleared for redistribution. STRF —operationalised-as→ TORCs STRF TORCs strength of the relation: 0.95 · note: RE-GROUNDS another row of this record: the old span named ONE endpoint. Quotation withheld because this record is not cleared for redistribution. ANOVA ANOVA Quotation withheld because this record is not cleared for redistribution. p-value p-value Quotation withheld because this record is not cleared for redistribution. median median Quotation withheld because this record is not cleared for redistribution. mean mean Quotation withheld because this record is not cleared for redistribution. ANOVA across SNR conditions ANOVA across SNR conditions Quotation withheld because this record is not cleared for redistribution. STRF power normalisation STRF power normalisation Quotation withheld because this record is not cleared for redistribution. an alternative noise-aware shape-change method an alternative noise-aware shape-change method Quotation withheld because this record is not cleared for redistribution. Engaging in the detection task reduced overall STRF gain in the majority of recorded cells relative to the passive state. Quotation withheld because this record is not cleared for redistribution. Shape changes, unlike gain changes, were excitatory and frequency-specific, concentrated in cells tuned near the target tone. Quotation withheld because this record is not cleared for redistribution. The combined gain and shape adaptations sharpened the target-versus-noise representation differently at high and low SNR. Quotation withheld because this record is not cleared for redistribution. The magnitude of receptive-field adaptation tracked behavioural performance, being greatest in the best-performing sessions. Quotation withheld because this record is not cleared for redistribution. Suppression of receptive fields grew as the task became harder and reached significance only in the low-SNR condition. Quotation withheld because this record is not cleared for redistribution. Behaviourally, lick withdrawal time lengthened as target SNR fell, in all three ferrets. Quotation withheld because this record is not cleared for redistribution. Identical intermediate probe trials were discriminated better within hard blocks than within easy blocks. Quotation withheld because this record is not cleared for redistribution. Naive animals exposed to the identical stimuli showed some minor but no consistent receptive-field changes. Quotation withheld because this record is not cleared for redistribution. The gain reduction was a broad, difficulty-independent effect present in near and far cells alike, in contrast to the difficulty-dependent shape changes. Quotation withheld because this record is not cleared for redistribution. Going from high to low SNR targets suppressed firing to the reference noise by roughly a third. Quotation withheld because this record is not cleared for redistribution. gain decreased in most cells during task performance while shape changes were excitatory and specific Quotation withheld because this record is not cleared for redistribution. grouped by distance from the target, near cells were on average enhanced and far cells suppressed Quotation withheld because this record is not cleared for redistribution. suppression became significant only in the low SNR tasks, at about -13% Quotation withheld because this record is not cleared for redistribution. the same tone placed far from a cell's BF produced suppression where near placement produced enhancement Quotation withheld because this record is not cleared for redistribution. the trend of increasing suppression at low SNR repeated in this subset, with the STRF changes amplified Quotation withheld because this record is not cleared for redistribution. naive animals given identical stimuli showed no consistent STRF changes Quotation withheld because this record is not cleared for redistribution. raising difficulty suppressed firing to the reference noise by about a third overall Quotation withheld because this record is not cleared for redistribution. probe trials of identical difficulty were detected about 10% better inside hard blocks than inside easy ones Quotation withheld because this record is not cleared for redistribution. gain was at or below 1 in most cells, with medians near 0.9 for near-high-SNR and 0.7 for far-low-SNR Quotation withheld because this record is not cleared for redistribution. contrast matched filter contrast matched filter Quotation withheld because this record is not cleared for redistribution. lick withdrawal time (LWT) lick withdrawal time (LWT) Quotation withheld because this record is not cleared for redistribution. TORC TORC Quotation withheld because this record is not cleared for redistribution. contrast matched filter contrast matched filter Quotation withheld because this record is not cleared for redistribution. lick withdrawal time lick withdrawal time Quotation withheld because this record is not cleared for redistribution. performance level performance level Quotation withheld because this record is not cleared for redistribution. STRF gain STRF gain Quotation withheld because this record is not cleared for redistribution. lick withdrawal time lick withdrawal time Quotation withheld because this record is not cleared for redistribution. naïve animals naïve animals Quotation withheld because this record is not cleared for redistribution. target SNR target SNR Quotation withheld because this record is not cleared for redistribution. discrimination rate discrimination rate Quotation withheld because this record is not cleared for redistribution. 237 — single cells in the population analysis of receptive-field plasticity, from two behaving ferrets Quotation withheld because this record is not cleared for redistribution. 93, 57, and 87 — single units in the high, mid and low SNR difficulty groups respectively Quotation withheld because this record is not cleared for redistribution. mean = −13.1 % — average amplitude change at best frequency in the low-SNR (hardest) condition Quotation withheld because this record is not cleared for redistribution. 10–30% — reduction in STRF gain caused by task engagement Quotation withheld because this record is not cleared for redistribution. 0.9 and 0.7 — median gain values for near cells in high-SNR tasks and far cells in low-SNR tasks (values below 1 mean suppression) Quotation withheld because this record is not cleared for redistribution. 350 ms — lick withdrawal time after target onset in the easiest (pure tone) detection task Quotation withheld because this record is not cleared for redistribution. 0.6 octaves — best-frequency distance criterion separating near cells from far cells Quotation withheld because this record is not cleared for redistribution. 85 — cells from the top-third best-performing sessions analysed separately Quotation withheld because this record is not cleared for redistribution. 93 units from 3 naïve animals — control sample recorded with identical stimuli but no behavioural task Quotation withheld because this record is not cleared for redistribution. 0.6 octaves — the near/far dividing line, called arbitrary and shown not to matter Quotation withheld because this record is not cleared for redistribution. 112 near versus 125 far cells — the balance the cut produced Quotation withheld because this record is not cleared for redistribution. 93, 57 and 87 single units in high, mid and low SNR — the difficulty groups Quotation withheld because this record is not cleared for redistribution. -13.1%, p < 0.05 — mean suppression, significant only at low SNR Quotation withheld because this record is not cleared for redistribution. 58.6% versus 69.2% — probe detection in easy versus hard blocks at identical SNR Quotation withheld because this record is not cleared for redistribution. 93 units from 3 naive animals — the no-behaviour control Quotation withheld because this record is not cleared for redistribution. 350 ms — lick withdrawal time for a pure-tone target 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.neuron.2008.12.027.