A question to carry into the reading
Does adding more context improve an answer when the relevant evidence moves to the middle?
Figure 1 and sections 2–3; record the models, tasks and position controls before discussing generality.
Further reading: MA-OPEN-028 · MA-OPEN-029
These are curatorial comparisons, not evidence supporting a scientific claim.
Reading and source
Results concern the tested historical models and task settings. They neither measure Mouseion nor establish the performance of current models; no benchmark was run.
The selected historical edition or paper is preserved in its original PDF. Consult the edition and transmission notes below.