kylelovesllms/grammar-v1-frames-ce-rb

Grammar V1 — CE-RB Populated Frames

The 40 depth-1 syntactic frames of grammar-v1-frames populated with 100 distinct randomly sampled lexicalizations per (frame, task) (seeded RNG, seed=0; grammar source: experiments/01_grammar_v1/src/cfg/grammar.py).

Partition (README.md / DataGenFlow.png): ambiguous (in-distribution) frames contribute copy and translate examples split 90/10 into train/validation (per frame and task, so every ambiguous frame appears in both); unambiguous (out-of-distribution) frames contribute copy examples to train and translate examples to test. Copy and translate rows of a frame never share a sentence, so no test translate input appears as a train copy input.

• train: 4960 rows = 1080 translate (ambiguous) + 1080 copy (ambiguous) + 2800 copy (unambiguous)
• validation: 240 rows = 120 translate + 120 copy (ambiguous only)
• test: 2800 rows (translate, unambiguous)

HF Sentence is the task target: the HI sentence itself for copy rows, the hierarchical (Head Final) translation for translate rows. Frames use indexed bracketed notation where [N_3] means the terminal filled by HI token 3.

Frame counts per category (before population)

Attachment Site	Gap Type	Frames	Ambiguous	Unambiguous
None	None	4	4	0
Subject	Subject Gap	16	0	16
Subject	Object Gap	8	0	8
Object	Subject Gap	8	8	0
Object	Object Gap	4	0	4
Total		40	12	28