Comment by iamflimflam1

Probably important to call out this part of the post:

Zero-shot vs. Fine-tuning: Out-of-the-box base models score ~0.35 on the typed-decisions benchmark (near random). The 0.766 score is achieved by fine-tuning on the benchmark's train split. Treat Laya as a fast foundation model to specialize, not as an omniscient zero-shot oracle.