---
title: dynaflip-base
canonical_url: "https://www.modelscope.cn/models/nextbig/dynaflip-base"
md_url: "https://www.modelscope.cn/models/nextbig/dynaflip-base.md"
repository: nextbig/dynaflip-base
last_updated: 2026-06-09
license: apache-2.0
model_type:
  - dynaflip
architectures:
  - DynaFLIPModel
parameters: 232.8M
tensor_type:
  - F32
library_name:
  - pytorch
  - transformer
  - safetensors
downloads: 9
stars: 0
---

# dynaflip-base

> dynaflip-base - nextbig 在 ModelScope 开源的模型。This model was proposed in DynaFLIP: Rethinking Robotics Perception via Tri-Modal-Dynamics Guided Representation.

nextbig/dynaflip-base 是 ModelScope 魔搭社区上的 232.8M 参数机器学习模型，采用 apache-2.0 许可。

- **Repository**: nextbig/dynaflip-base
- **License**: apache-2.0
- **Parameters**: 232.8M
- **Downloads**: 9
- **Stars**: 0
- **Last updated**: 2026-06-09

Source: https://www.modelscope.cn/models/nextbig/dynaflip-base

---

# DynaFLIP-base

This model was proposed in [DynaFLIP: Rethinking Robotics Perception via Tri-Modal-Dynamics Guided Representation](https://huggingface.co/papers/2605.30350).

## Usage

The model is compatible with Transformers:

```python
from transformers import AutoModel, AutoProcessor
from PIL import Image
import torch

REPO = "jlee-larr/dynaflip-base"
dynaflip = AutoModel.from_pretrained(REPO, trust_remote_code=True).eval()
processor = AutoProcessor.from_pretrained(REPO, trust_remote_code=True)

image = Image.open("example.png").convert("RGB")
inputs = processor(images=image, return_tensors="pt")

with torch.no_grad():
    v = dynaflip.vision_outputs(inputs["pixel_values"])

# v.last_hidden_state  -> (B, num_patches, 768)   patch tokens
# v.pooler_output      -> (B, 1536)               CLS + mean(patches)
```

## Citation

```bibtex
@article{lee2026dynaflip,
  title   = {DynaFLIP: Rethinking Robotics Perception via Tri-Modal-Dynamics Guided Representation},
  author  = {Lee, Jusuk and Lee, Seungjae and Shin, Jonghun and Jung, Hoseong and Kim, Sungha and Cho, Daesol and Kim, H. Jin and Huang, Jia-Bin and Huang, Furong},
  journal = {arXiv preprint arXiv:2605.30350},
  year    = {2026},
}
```