Stimulus-specific regularities as a basis for perceptual induction
Bibliographic record
Abstract
A hallmark of visual intelligence is the ability to extract relationships among objects. One form of extraction produces stimulus-specific knowledge (statistical learning). Another form produces stimulus-general principles (inductive learning). These two learning processes seem incompatible on the surface, but may be related on a deeper level. Here we examine how statistical learning and inductive learning interact. In Experiment 1, observers were randomly assigned to one of three conditions where they viewed a sequence of circles of varying sizes. In the rule+regularities condition, the sequence contained repeated triplets (regularities) that followed a rule: the three circles increased in size in each triplet. In the rule-only condition, the sequence contained sets of three circles that followed the same rule, but all sets were unique (no regularities). In the regularities-only condition, the sequence contained repeated triplets that did not follow the rule. We found that learning of the rule and learning of the regularities were both stronger in the rule+regularities condition than in the rule-only, or the regularities-only condition. This suggests that statistical learning and inductive learning are mutually beneficial. To tease apart whether one learning process is necessary for the other, we increased the complexity of the rule in Experiment 2. Everything was identical to Experiment 1, except now the rule was that within each triplet or set, the first circle was smaller than the second circle, and the second was larger than the third. Learning of the rule was only successful in the rule+regularities condition, but not in the rule-only condition. Moreover, there was no learning of regularities. This suggests that the presence of regularities facilitated rule learning, but the presence of the rule did not help the learning of regularities. These findings demonstrate that statistical learning and inductive learning are related, and that stimulus-specific regularities are necessary for inductive learning. Meeting abstract presented at VSS 2015
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 | 0.002 | 0.013 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
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".