Leveraging computational and animal models of vision to probe atypical emotion recognition in autism
Bibliographic record
Abstract
Recognizing others' emotions based on facial expressions is a core component of human social interactions. Previous studies (Wang and Adolphs 2017) have suggested that autistic individuals show differences in their facial emotion recognition compared to neurotypical adults. What are the neural mechanisms that account for these observed differences? Here we lay the groundwork for a new approach combining cutting-edge computational and empirical non-human primate work to test theories of atypical facial emotion recognition in autistic adults. In a recent study, the author(s) observed that artificial neural network (ANN) models of vision developed to achieve a myriad of visual objectives (e.g., object, emotion and face identification) could be fine-tuned to perform facial emotion judgments. Interestingly, the ANNs' image-level behavioral patterns better matched the neurotypical subjects' compared to autistic adults. This behavioral mismatch was most remarkable when the ANN behavior was constructed from units that correspond to the primate inferior temporal (IT) cortex. Here we directly test these two predictions in the rhesus macaques. First, we trained two macaques to perform a binary facial emotion (happy vs. fearful) discrimination task. Consistent with ANN predictions, the macaque image-level behavioral patterns better matched the behavior obtained in human Controls than in autistic individuals. Second, we implanted multi-electrode arrays in the IT cortex of two macaques and performed large-scale neural recordings while they fixated on images (used in the Wang and Adolphs study). Using the recorded neural multiunit spiking activity, we built regression models (165 IT-based models tested) to predict facial emotion ground truth ("level of happiness") on held-out images. Consistent with ANN-IT predictions, macaque IT population decodes of facial emotions better matched the neurotypical behavior compared to autistic individuals. Our results, therefore, establish the rhesus macaque as an appropriate species to further probe the neurobehavioral markers with ANN-guided hypotheses and experiment design.
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) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| 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 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".