Temporal evolution of colour representation measured with magnetoencephalography (MEG).
Bibliographic record
Abstract
Colour perception is based on the differential spectral responses of the L,M and S-cones, subsequent subcortical and cortical computations, and may include the influence of higher order factors such as language. Due to our much higher sensitivity to L/M cone opponent contrast than S-cone contrast, a colour space that is ‘perceptually uniform’ is very different to one defined by cone contrast. Here we compare the evolution of different colour representations over time using magnetoencephalography (MEG). We measured neural responses to 14 hues at each of 3 achromatic offsets (increment, equiluminant and decrement). Stimuli were circular blobs (40-deg diameter) of 300ms duration, smoothed in space and time, each presented 42 times in a counterbalanced order. For each subject (n=8), we trained classifiers to discriminate all possible pairs of stimuli using MEG responses at each timepoint (10ms bins). From ~100ms after stimulus-onset, we found robust classification of stimuli varying in hue and/or achromatic offset. For stimuli differing only in achromatic offset, classifier performance peaked at ~160ms after stimulus onset, then decayed. For stimuli that varied only in hue, classifier performance remained high ~160-400ms. This suggests that while early neural responses differentiate stimuli both on chromatic and achromatic content, they then shift towards a more colour-dominated response. Using Representational Similarity Analysis (RSA) we found that models based on colour, including hue and colour category, correlated well with the data at later times (after 400ms). Surprisingly, these models did not perform significantly better than one based on raw cone contrast, even though cone contrast is not a good predictor of perceptual measures such as visibility and discriminability. We conclude that to demonstrate the emergence of neural representations of perceived hue or colour category, it is crucial to demonstrate that these ‘higher-level’ models provide a better account than lower-level stimulus representations, such as cone contrast.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".