English

SAGE-32B: Agentic Reasoning via Iterative Distillation

Artificial Intelligence 2026-04-22 v2 Computation and Language Machine Learning

Abstract

We demonstrate SAGE-32B, a 32 billion parameter language model that focuses on agentic reasoning and long range planning tasks. Unlike chat models that aim for general conversation fluency, SAGE-32B is designed to operate in an agentic loop, emphasizing task decomposition, tool usage, and error recovery. The model is initialized from the Qwen2.5-32B pretrained model and fine tuned using Iterative Distillation, a two stage training process that improves reasoning performance through rigorously tested feedback loops. SAGE-32B also introduces an inverse reasoning approach, which uses a meta cognition head to forecast potential failures in the planning process before execution. On agentic reasoning benchmarks including MMLU-Pro, AgentBench, and MATH-500, SAGE-32B achieves higher success rates in multi tool usage scenarios compared to similarly sized baseline models, while remaining competitive on standard reasoning evaluations. Model weights are publicly released at https://huggingface.co/sagea-ai/sage-reasoning-32b

Keywords

Cite

@article{arxiv.2601.04237,
  title  = {SAGE-32B: Agentic Reasoning via Iterative Distillation},
  author = {Basab Jha and Firoj Paudel and Ujjwal Puri and Ethan Henkel and Zhang Yuting and Mateusz Kowalczyk and Mei Huang and Choi Donghyuk and Wang Junhao},
  journal= {arXiv preprint arXiv:2601.04237},
  year   = {2026}
}

Comments

23 Pages, 3 figures, 4 tables

R2 v1 2026-07-01T08:54:54.798Z