ROM: Real-time Overthinking Mitigation via Streaming Detection and Intervention

Xinyan Wang¹, Xiaogeng Liu², Chaowei Xiao²

¹University of Wisconsin-Madison   ²Johns Hopkins University

Overview

ROM is a lightweight streaming framework that detects and mitigates overthinking in Large Reasoning Models (LRMs) at the token level. It attaches a small detection head (< 0.01% additional parameters) to a frozen LLM backbone and monitors generation in real time, intervening when redundant reasoning is detected.

Method

How ROM Works: A GSM8K Example

Project Structure

ROM/
├── rom/                        # Core package
│   ├── models.py               # StreamingHead, Qwen3WithHead
│   ├── dataset.py              # Dataset loading & embedding cache
│   ├── train.py                # Training pipeline
│   ├── eval.py                 # Offline evaluation (vLLM)
│   ├── env.py                  # Environment setup
│   └── utils/
│       ├── math.py             # Answer extraction & correctness checking
│       └── eval_helpers.py     # Metrics, probability computation
├── configs/
│   ├── train.yaml              # Training defaults
│   └── eval.yaml               # Evaluation defaults
├── requirements.txt
├── LICENSE
└── README.md

Quick Start

Installation

pip install -r requirements.txt

Requires Python 3.11+, PyTorch >= 2.9.0, and a CUDA-capable GPU.

Data

Training data is hosted on HuggingFace: xinyan-wang/ROM.

Download and place under data/:

# Using huggingface-cli
huggingface-cli download xinyan-wang/ROM --repo-type dataset --local-dir data

Training

All parameters are in configs/train.yaml. Run with defaults:

python -m rom.train

Override via CLI:

python -m rom.train --lr 1e-4 --num_train_epochs 30

W&B logging is enabled by default. Disable with --no_wandb.

Evaluation

python -m rom.eval

Override as needed:

python -m rom.eval --ckpt_path checkpoints/my_model.pt --test_data data/test_data/math500.jsonl

Citation

If you find ROM useful, please cite our paper 📝 and give us a ⭐!

@article{wang2025rom,
  title={ROM: Real-time Overthinking Mitigation via Streaming Detection and Intervention},
  author={Wang, Xinyan and Liu, Xiaogeng and Xiao, Chaowei},
  journal={arXiv preprint arXiv:2603.22016},
  year={2025}
}

License

This project is licensed under the MIT License.