What the temporal dynamics of unimodal sensory estimation can tell us about statistically optimal multimodal integration
Bibliographic record
Abstract
The brain combines noisy, redundant stimulus estimates (e.g., position of an object provided by vision and haptics) in proportion to their reliabilities. The mechanisms used by the brain to assess reliability (inverse of noise) of its estimates are, however, not fully understood. Most models make the untested assumption that internal estimates of a property respond rapidly to fluctuations arising from noise. We tested this assumption by employing a visual stimulus consisting of an array of flickering vertical lines distributed to produce an area of higher density which the subject had to track. This stimulus was similar to one that has been shown (Byrne and Henriques, 2013) to combine optimally with haptic estimates of position. The area of higher density was inconspicuously jumped left or right at varying time intervals before the subject indicated its location relative to a reference line. By looking for whether the jump affected subjects’ judgements, we could assess the time required for updating internal position estimates. Subjects’ responses showed that large jumps had little effect on position estimates until the post-jump stimulus was present for over 300 ms. At this time the estimate suddenly transitioned to a new location. Updating for smaller jumps proceeded more smoothly. These results imply that position estimates are maintained in an ‘attractor’, which takes time to shift when sensory information changes. This is inconsistent with current models of how reliability is computed by the brain because such models, which rely on population coding or temporal sampling, require estimates to fluctuate in lock-step with noise.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.009 | 0.002 |
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; both teacher heads agree on what is shown here.
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".