laguna-s

A user notices that the Q4_K_M quantized version of Laguna S 2.1 increased from 68GB to 96GB, likely due to using more FP16 layers, and discusses potential issues with quantization and context looping.

Explores whether using quantization lower than Q8 degrades reasoning in Laguna S 2.1, causing it to seem 'stupid'.

A detailed comparison of the new open-weights Laguna-S-2.1 mixture-of-experts model against Qwen 3.6-35B-A3B, highlighting how Laguna's larger active parameter count and long-horizon focus make it superior for extended tasks despite slower speed.

Poolside releases Laguna S 2.1, a 118B parameter Mixture-of-Experts model with 8B activated per token and 1M context window, claiming to beat models three times its size like DeepSeek v4 Pro, Gemini 3.6 Flash, and Thinking Machines Inkling.