Base Labs Says mammoth-1 Tops 208 Decision Models on Jev Index
Summary
Base Labs releases mammoth-1 and elephant-1, two decision models built by fine-tuning Qwen3.5-27B and Qwen3.5-9B on 6.21 million examples, with weights published. Base Labs says mammoth-1 ranks first of 208 models on the Jev Decision Index 0.4 at 71.24, 1.8 points ahead of Perplexity Decider v1.2, and that elephant-1 scores 62.65, best under 15B parameters.
Key Points
- Base Labs trains both models with a block-bidirectional attention mask on 16 full-attention layers, scoring 48.6 against 44.8 for causal attention.
- Base Labs says elephant-1 ties Liquid AI's d1-ultra at 65 on the Artificial Analysis Decision Index.
- Mammoth-1 trains on 18 nodes of four GB300 GPUs each, 72 GPUs in total.