Tagged: GPT-OSS

4 articles

Latest

4-bit GPT-OSS 60B beats its bfloat16 original on 7 of 9 benchmarks

OpenAI's worst-case fine-tuning test found GPT-OSS still trails o3 on dangerous capabilities

GPT-OSS training blew up until LinkedIn engineers fixed a sneaky MoE routing bug

OpenAI drops GPT-OSS: a 120B open-source MoE model that fits on a single 80GB GPU