Perception is subjective. Even basic judgments, like those of visual object size, vary substantially between observers and also across the visual field within the same observer. The way in which the visual system determines the size of objects remains unclear, however. We hypothesize that object size is inferred from neuronal population activity in V1 and predict that idiosyncrasies in cortical functional architecture should therefore explain individual differences in size judgments. Here we show results from novel behavioural methods and functional magnetic resonance imaging (fMRI) demonstrating that biases in size perception are correlated with the spatial tuning of neuronal populations in healthy volunteers. To explain this relationship, we formulate a population read-out model that directly links the spatial distribution of V1 representations to our perceptual experience of visual size. Taken together, our results suggest that the individual perception of simple stimuli is warped by idiosyncrasies in visual cortical organization.

14 Perception is subjective. Even basic judgments, like those of visual object size, vary substantiallyHow do we perceive the size of an object? A range of recent observations have lent support to the 26 hypothesis that the visual system generates the perceived size of an object from its cortical 27 representation in early visual cortex 1 . In particular, the spatial spread of neural activity in visual 28 cortex has been related to apparent size under a range of contextual modulations [2][3][4][5][6][7] . The strength of 29 contextual size illusions has further been linked to the cortical territory in V1 that represents the 30 central visual field 8,9 . These findings suggest that lateral connections in V1 may play a central role in 31 size judgments because these interactions are reduced when V1 surface area is larger. Indeed, 32 similar interactions have been argued to underlie the strength of the tilt illusion 10,11 , perceptual 33 alternations in binocular rivalry 12 , the influence of distractors in visual search tasks 13 , and visual 34 working memory capacity 14 . Even the precision of mental imagery co-varies with V1 area 15 35 suggesting V1 may be used as a 'workspace' for storing mental images whose resolution is better 36 when surface area is larger. 37. CC-BY-NC 4.0 International license not peer-reviewed) is the author/funder. It is made available under a The copyright holder for this preprint (which was . http://dx.doi.org/10.1101/026989 doi: bioRxiv preprint first posted online Sep. 16, 2015; 2 However, these previous findings do not demonstrate that V1 representations per se are relevant for 38 size judgments, and in particular for subjective judgments of object size. If V1 signals were indeed 39 the basis for these judgments then variations in the functional architecture of V1 should explain 40 idiosyncratic biases in basic size perception (i.e. size judgements that occur in the absence of any 41 contextual/illusory effects). To date this prediction remains untested. Previous neuroimaging 42 experiments have focused on modulations of apparent size that must involve additional processing, 43 either due to local interactions between adjacent stimuli in V1 or by a context that likely involves 44 processing in higher visual areas. Others have shown that the objective ability to discriminate subtle 45 differences between stimuli is related to cortical magnification and spatial tuning in early visual 46 cortex 11,16,17 . However, no experiment to date has shown a relationship between V1 and subjective 47 perceptual biases in the absence of any contextual interaction, even though there are considerable 48 individual differences in perceptual biases. 49It is well established that subjective size judgments for simple, small stimuli can vary substantially 50 between observers and even across the visual field within the same observer. Previous behavioral 51 research has shown that small visual stimuli appear smaller when they are presented in the 52 periphery [18][19][20] . A ...

There are 4 cone morphologies in zebrafish, corresponding to UV (U), blue (B), green (G) and red (R)-sensing types, yet genetically, 8 cone opsins are expressed. How 8 opsins are physiologically siloed in 4 cone types is not well understood, and in larvae, cone physiological spectral peaks are unstudied. We use a spectral model to infer cone wavelength peaks, semi-saturation irradiances, and saturation amplitudes from ERG datasets composed of multi-wavelength, multi-irradiance, Aspartate-isolated, cone-PIII signals, as compiled from many 5–12-day larval, and 8–18-month adult eyes isolated from WT or roy orbeson (roy) strains. Analysis suggests (in nm) a 7-cone, U-360 / B1–427 / B2–440 / G1–460 / G3–476 /R1–575 / R2–556 spectral physiology in WT larvae, but a 6-cone, U-349 / B1–414 / G3–483 / G4–495 / R1–572 / R2–556 structure in WT adults. In the roy larvae there is a 5-cone structure: U-373 / B2–440 / G1–460 / R1–575 / R2–556; in roy adults, a 4-cone structure, B1–410 / G3–482 / R1–571 / R2–556. Existence of multiple B, G, and R types is inferred from shifts in peaks with red or blue backgrounds. Cones were either high or low semi-saturation types. The more sensitive, low semi-saturation types included U, B1 and G1 cones [3.0–3.6 log(quanta·μm−2·s−1)]. The less sensitive, high semi-saturation types were B2, G3, G4, R1 and R2 types [4.3–4.7 log(quanta·μm−2·s−1)]. In both WT and roy, U- and B- cone saturation amplitudes were greater in larvae than adults, while G-cone saturation levels were greater in adults. R-cone maximal amplitudes were the largest (50–60% of maximal dataset amplitudes), and constant throughout development. WT and roy larvae differed in cone sisgnal levels, with lesser UV- and greater G-cone amplitudes occurring in roy, indicating strain variation in physiological development of cone signals. These physiological measures of cone types suggest chromatic processing in zebrafish involves at least 4–7 spectral-signal processing pools.

Perceptual bias is inherent to all our senses, particularly in the form of visual illusions and aftereffects. However, many experiments measuring perceptual biases may be susceptible to nonperceptual factors, such as response bias and decision criteria. Here, we quantify how robust multiple alternative perceptual search (MAPS) is for disentangling estimates of perceptual biases from these confounding factors. First, our results show that while there are considerable response biases in our four-alternative forcedchoice design, these are unrelated to perceptual biases estimates, and these response biases are not produced by the response modality (keyboard vs. mouse). We also show that perceptual bias estimates are reduced when feedback is given on each trial, likely due to feedback enabling observers to partially (and actively) correct for perceptual biases. However, this does not impact the reliability with which MAPS detects the presence of perceptual biases. Finally, our results show that MAPS can detect actual perceptual biases and is not a decisional bias towards choosing the target in the middle of the candidate stimulus distribution. In summary, researchers conducting a MAPS experiment should use a constant reference stimulus, but consider varying the mean of the candidate distribution. Ideally, they should not employ trial-wise feedback if the magnitude of perceptual biases is of interest.

The developmental progression of eight opsin spectral signals recorded from the zebrafish retinal cone layer is altered by the timing and cell type expression of thyroxin receptor β2 (trβ2) gain-offunction transgenes.