Use of a correlative training method in the rehabilitation of acquired prosopagnosia
Bibliographic record
Abstract
No effective treatment is known for acquired prosopagnosia. We investigated a novel rehabilitative training strategy, based on work with neural network models showing that correlating a weak cue with a strong cue during training can help a network learn tasks that would otherwise not be possible. Many prosopagnosic subjects can recognize facial expressions despite their problems with identity. By correlating expression with identity during early training stages, we can pair a strong cue (expression) with a weak one (identity). With repeated training, this correlation should increase the perceived difference between these novel faces, eventually allowing recognition of identity even when expression is no longer correlated. We trained two prosopagnosic subjects (R-AT1 and B-AT1) with anterior temporal lesions and intact recognition of facial expression. During the correlative method, subjects learned five frontal-view faces, initially all with unique expressions. Once they achieved a criterion success rate, a modest degree of variability in expression was introduced, and more again once criterion was achieved, until expression was eventually uncorrelated with identity after several weeks of thrice-weekly training. Additional training runs were performed with hair removed, and external contour removed. As control experiments, we had subjects learn five other faces over a similar time period, but without any correlation between identity and expression. Subjects learned to recognize these small sets of faces, even without hair or external contour, and showed high levels of retention even two months later. However, subjects also learned the faces in control experiments, suggesting that repeated exposure was also effective. fMRI scanning in one subject showed a significant increase in peak-voxel significance and the number of face-selective voxels in the fusiform face area after training. These results show that prosopagnosics can learn to recognize a small set of faces with at least some invariance for expression.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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 source (direct Gemma or distilled Codex), 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".