---
base_model: Qwen/Qwen3-4B
language:
- en
library_name: transformers
license: apache-2.0
pipeline_tag: text-generation
tags:
- metacognitive-behavioral-tuning
- multi-hop-qa
- reasoning
- sft
- grpo
- MBT-R
---

# Qwen3-4B · MBT-R

MBT-R (main table) — **final checkpoint (SFT → GRPO)**. Base: [`Qwen/Qwen3-4B`](https://huggingface.co/Qwen/Qwen3-4B).
Paper: *Metacognitive Behavioral Tuning of Large Language Models for Multi-Hop Question Answering*.

- **Method**: MBT-R (Refinement): the student's own reasoning traces are rewritten into the 5-phase structure for SFT, then GRPO.
- **Base model**: `Qwen/Qwen3-4B`
- **Training**: SFT (LR 1e-4, BS 128, HotpotQA) → GRPO
- **Benchmarks**: HotpotQA (ID), MuSiQue / 2WikiMultiHopQA (OOD)

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "metacognitive-behavioral-tuning/Qwen3-4B-MBT-R"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo, dtype="bfloat16", device_map="auto")
```