AnalysisAI Models
29 days ago
Gemma-4-Gembrain 31B merge released on Hugging Face
Community merge of multiple Gemma 4 31B instruction finetunes, aiming to improve logical and lateral thinking. Features KLD of 0.0186 and only 13 refusals out of 100.
Community merge of multiple Gemma 4 31B instruction finetunes, aiming to improve logical and lateral thinking. Features KLD of 0.0186 and only 13 refusals out of 100.