Provides six open language models from 0.9B to 375B-A23B for reasoning and agentic tasks. The paper also introduces sparse attention and a decoding speedup that needs no separate draft model.
A LoRA adapter that tunes Qwen3.8-27B for software-engineering and terminal work in the pi coding agent. Its creators report higher pass rates and fewer tool calls on selected benchmarks.
Reuses decoder-layer weights to increase computation depth at roughly 9B parameters, taking text and images as input and generating text with up to 512K-token context. It targets reasoning, coding, and tool use, though performance on complex long-horizon tasks can be inconsistent.