Readability and difficulty-controlled text generation on OSE benchmark leaderboard · SOTA2 Research

Readability and difficulty-controlled text generation on OSE

0.8453AccuracySWAI

Evaluation Results

Method	Links
SWAI2026.01		0.8453	0.845	0.8453	0.8453	0.847	0.768	0.768
SWAI2026.01		0.7111	0.712	0.7127	0.7111	0.805	0.567	0.567
Baseline2026.01		0.3711	0.363	0.362	0.3648	0.846	0.052	0.052
Baseline2026.01		0.2887	0.289	0.2936	0.2899	0.826	-0.063	-0.064