BabyVision: Results as Published

BabyVision tests core visual abilities that do not depend on language: basic visual tasks that young children solve, grouped into four categories (fine-grained discrimination, visual tracking, spatial perception and visual pattern recognition) across 388 items in 22 subclasses. Scores are accuracy against human baselines; the paper reports the best model at 49.7 against an adult average of 94.1.

Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.