Intracranial Entrainment Reveals Statistical Learning across Levels of Abstraction
Bibliographic record
Abstract
We encounter the same people, places, and objects in predictable sequences and configurations. Humans efficiently learn these regularities via statistical learning. Importantly, statistical learning creates knowledge not only of specific regularities but also of regularities that apply more generally across related experiences (i.e., across members of a category). Prior evidence for different levels of learning comes from post-exposure behavioral tests, leaving open the question of whether more abstract regularities are detected online during initial exposure. We address this question by measuring neural entrainment in intracranial recordings. Neurosurgical patients viewed a stream of photographs with regularities at one of two levels: In the exemplar-level structured condition, the same photographs appeared repeatedly in pairs. In the category-level structured condition, the photographs were trial-unique but their categories were paired across repetitions. In a baseline random condition, the same photographs repeated but in a scrambled order. We measured entrainment at the frequency of individual photographs, which was expected in all conditions, but critically also at half that frequency-the rate at which to-be-learned pairs appeared in the two structured (but not random) conditions. Entrainment to both exemplar and category pairs emerged within minutes throughout visual cortex and in frontal and temporal regions. Many electrode contacts were sensitive to only one level of structure, but a significant number encoded both levels. These findings suggest that the brain spontaneously uncovers category-level regularities during statistical learning, providing insight into the brain's unsupervised mechanisms for building flexible and robust knowledge that generalizes across input variation and conceptual hierarchies.
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 | 0.009 |
| 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 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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".