---
language:
- en
license: apache-2.0
library_name: transformers
base_model: Qwen/Qwen3-1.7B
tags:
- agent
- tool-use
- distillation
- math
- code
- reasoning
pipeline_tag: text-generation
---

<div align="center">

<h1>SOD-1.7B</h1>

<p>
  <a href="https://arxiv.org/abs/2605.07725">
    <img src="https://img.shields.io/badge/Paper-Arxiv-red?logo=arxiv&logoColor=red" alt="Paper on arXiv"/>
  </a>
  <a href="https://github.com/YoungZ365/SOD">
    <img src="https://img.shields.io/badge/Code-GitHub-black?logo=github&logoColor=white" alt="Code on GitHub"/>
  </a>
  <a href="https://huggingface.co/collections/youngzhong/sod-6a03530369d76913c24a4ffb">
    <img src="https://img.shields.io/badge/Collection-SOD-yellow?logo=huggingface" alt="HuggingFace Collection"/>
  </a>
</p>

</div>

## About

**SOD-1.7B** is a 1.7B student model distilled from a 4B teacher using **SOD (Step-wise On-policy Distillation)**, a method designed for training small language model agents with tool-integrated reasoning capabilities.

SOD addresses the **cascading error propagation** problem in on-policy distillation for agentic reasoning by introducing an adaptive step-level weighting mechanism that suppresses distillation loss on drifted steps and restores supervision when the student recovers alignment — all at negligible additional computational cost.

## Model Information

| Attribute | Value |
|-----------|-------|
| Base Model | [Qwen3-1.7B](https://huggingface.co/Qwen/Qwen3-1.7B) |
| Teacher Model | [SOD-GRPO_teacher-4B](https://huggingface.co/youngzhong/SOD-GRPO_teacher-4B) |
| Training Pipeline | Cold-Start SFT → SOD (Step-wise On-policy Distillation) |
| Parameters | 1.7B |

## Related Models

| Model | Description |
|-------|-------------|
| [SOD-0.6B](https://huggingface.co/youngzhong/SOD-0.6B) | SOD-distilled 0.6B student |
| [SOD-1.7B](https://huggingface.co/youngzhong/SOD-1.7B) | SOD-distilled 1.7B student (this model) |
| [SOD-GRPO_teacher-4B](https://huggingface.co/youngzhong/SOD-GRPO_teacher-4B) | GRPO-trained 4B teacher model |

## Performance

We report **average@32** over 5 runs on challenging math, science, and code benchmarks.

### 1.7B Student Results

| Method | AIME 2024 | AIME 2025 | GPQA-Diamond | LiveCodeBench-v6 | Average |
|--------|-----------|-----------|--------------|------------------|---------|
| Vanilla | 9.90 | 8.96 | 26.80 | 22.73 | 17.10 |
| SFT | 26.77 | 22.40 | 29.85 | 24.63 | 25.91 |
| GRPO | 25.63 | 21.67 | 33.55 | 20.70 | 25.39 |
| OPD | 43.86 | 37.04 | 31.73 | 32.45 | 36.27 |
| OPSD_gt | 33.85 | 24.69 | 35.02 | 22.73 | 29.07 |
| OPSD_hint | 34.42 | 21.43 | 33.46 | 23.12 | 28.11 |
| **SOD (This Model)** | **50.83** | **41.72** | **38.72** | **40.63** | **42.98** |

### Teacher Model (4B)

| Method | AIME 2024 | AIME 2025 | GPQA-Diamond | LiveCodeBench-v6 | Average |
|--------|-----------|-----------|--------------|------------------|---------|
| GRPO | 67.60 | 60.42 | 55.19 | 63.13 | 61.59 |

## Key Highlights

- 🏆 **Recovers 69.8% of teacher performance** with only 1.7B parameters (42.98 vs 61.59)
- 📈 **+18.5% over second-best baseline** (OPD) on average
- 💡 **Minimal extra compute**: The divergence metric reuses log-probabilities already computed in the forward pass

## Citation

```bibtex
@article{zhong2026sod,
      title={SOD: Step-wise On-policy Distillation for Small Language Model Agents}, 
      author={Qiyong Zhong and Mao Zheng and Mingyang Song and Xin Lin and Jie Sun and Houcheng Jiang and Xiang Wang and Junfeng Fang},
      journal={arXiv preprint arXiv:2605.07725},
      year={2026}
}
```