--- license: apache-2.0 language: - en pipeline_tag: any-to-any tags: - image-generation - interleaved-generation - multimodal-reasoning - visual-reasoning - vbvr-pro base_model: ThinkMorph/ThinkMorph-7B base_model_relation: finetune --- # VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning

## Overview Native visual reasoning, i.e., reasoning through visual generation, has recently emerged as a promising direction for studying visual intelligence beyond language. Yet progress remains bottlenecked by the lack of scalable training tasks, reliable feedback, and controlled comparisons across generative substrates. In this work, we introduce **VBVR-Pro**, a closed-loop testbed that makes native visual reasoning through generation trainable, verifiable, optimizable, and experimentally controllable. **1) Task scaling.** VBVR-Pro turns visual reasoning into a controlled task space of *300* procedurally generated tasks. Models trained on VBVR-Pro show strong transfer beyond the proposed suite across *six* held-out visual reasoning benchmarks such as RISE-Video, MME-CoF-Pro, and BabyVision. Further analysis validates that these gains reflect visual reasoning rather than instruction-pattern fitting. **2) Verifiable rewards.** VBVR-Pro provides verifiable reward scorers for task-grounded evaluation. Through a systematic study of leading MLLMs as judges, we identify recurring failure modes of the prevalent *VLM-as-a-judge* paradigm. In contrast, the proposed scorers are grounded on verifiable task-specific rules, achieve fine-grained alignment with human judgments. Importantly, they serve as reliable reward signals for large-scale multi-task reinforcement learning and demonstrate stronger post-RL performance across visual reasoning tasks. **3) Mechanism study.** VBVR-Pro enables controlled modality studies across more than *30* image, video, and interleaved generators. Our analysis shows that video generation remains strongest for tasks requiring persistent spatiotemporal state tracking, while interleaved generation provides a compute-efficient alternative by externalizing intermediate visual states. Critically, ablations and probing confirm the presence of vision-native trajectories, that are a more crucial substrate than explicit linguistic chains of thought for visual reasoning. We release all data, models, scorers, and code to facilitate future research. The models are presented in the paper [VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning](https://huggingface.co/papers/2602.20159). ## Models Zoo

Model	Base Architecture	Other Remarks
Image Generation Models
VBVR-Pro-BAGEL	BAGEL-7B-MoT	Complete model
VBVR-Pro-FLUX2-dev	FLUX.2-dev	Complete model, Diffusers format
VBVR-Pro-FLUX2-dev-diffsynth	FLUX.2-dev	LoRA model, DiffSynth format
VBVR-Pro-Qwen-Image-Edit	Qwen-Image-Edit-2511	Complete model, Diffusers format
VBVR-Pro-Qwen-Image-Edit-diffsynth	Qwen-Image-Edit-2511	LoRA model, DiffSynth format
Interleaved Image Generation Models
VBVR-Pro-ThinkMorph	ThinkMorph-7B	Complete model
VBVR-Pro-SenseNova-U1	SenseNova-U1-8B-MoT	Complete model
Video Generation Models
VBVR-Pro-LTX2.3	LTX-Video-2.3	Complete model, Diffusers format
VBVR-Pro-LTX2.3-diffsynth	LTX-Video-2.3	LoRA model, DiffSynth format
VBVR-Pro-Wan2.1-I2V-14B	Wan2.1-I2V-14B-720P	Complete model, Diffusers format
VBVR-Pro-Wan2.1-I2V-14B-diffsynth	Wan2.1-I2V-14B-720P	LoRA model, DiffSynth format
VBVR-Pro-Wan2.2-I2V-A14B	Wan2.2-I2V-A14B	Complete model, Diffusers format
VBVR-Pro-Wan2.2-I2V-A14B-diffsynth	Wan2.2-I2V-A14B	LoRA model, DiffSynth format
VBVR-Pro-Wan2.2-TI2V-5B	Wan2.2-TI2V-5B	Complete model, Diffusers format
VBVR-Pro-Wan2.2-TI2V-5B-diffsynth	Wan2.2-TI2V-5B	LoRA model, DiffSynth format

## Release Information This repository contains a VBVR-Pro fine-tune of [ThinkMorph-7B](https://huggingface.co/ThinkMorph/ThinkMorph-7B). It takes an initial image and a text instruction, reasons with interleaved text and images, and can generate one or more sequential images before returning its final answer. The released weights are the EMA checkpoint from training step 25,000. The training checkpoint was stored in FP32; this Hugging Face export is converted to BF16, matching the precision of the original ThinkMorph-7B release. ### Repository contents | File | Purpose | | --- | --- | | `ema.safetensors` | Complete BF16 fine-tuned ThinkMorph model state | | `ae.safetensors` | Image autoencoder used for encoding and generation | | `llm_config.json` | Qwen2-MoT language-model configuration | | `vit_config.json` | SigLIP vision-encoder configuration | | `tokenizer.json`, `tokenizer_config.json`, `vocab.json`, `merges.txt` | Tokenizer assets | | `generation_config.json` | Default text-generation settings | | `config.json` | Checkpoint family identifier | This is a full checkpoint, not a LoRA or other adapter. No separate base-model download is needed after this repository has been downloaded. In this release, we present all models presented in paper [**VBVR-Pro-Trained-Models**](https://huggingface.co/collections/Video-Reason/VBVR-Pro), [**VBVR-Pro-Dataset-Video**](https://huggingface.co/datasets/Video-Reason/VBVR-Pro-SFT-Video), [**VBVR-Pro-Dataset-Image**](https://huggingface.co/datasets/Video-Reason/VBVR-Pro-SFT-Image), [**VBVR-Pro-Bench**](https://huggingface.co/datasets/Video-Reason/VBVR-Pro-Bench), [**VBVR-Pro-Code**](https://github.com/Video-Reason/VBVR-Pro) and [**VBVR-Bench-Leaderboard**](https://video-reason.com/pro/bench/#leaderboard). ## VBVR-Pro Benchmark Results

Models	Overall	In-Domain by Category						Out-of-Domain by Category
Models	Overall	Avg.	Abst.	Know.	Perc.	Spat.	Trans.	Avg.	Abst.	Know.	Perc.	Spat.	Trans.
Image Generation Models
Proprietary Models
Qwen-Image-2.0	0.313	0.248	0.269	0.196	0.225	0.170	0.132	0.378	0.341	0.235	0.391	0.384	0.080
Seedream-5.0-Pro	0.557	0.485	0.518	0.312	0.509	0.401	0.217	0.629	0.507	0.455	0.661	0.559	0.202
Open-source Models
BAGEL-7B-MoT	0.089	0.066	0.039	0.085	0.067	0.046	0.027	0.111	0.201	0.031	0.073	0.028	0.121
FLUX.2-dev	0.157	0.108	0.088	0.109	0.072	0.100	0.066	0.206	0.197	0.165	0.184	0.241	0.077
Qwen-Image-Edit	0.134	0.108	0.092	0.082	0.100	0.109	0.056	0.159	0.176	0.063	0.141	0.182	0.082
Strong Baselines
VBVR-Pro-BAGEL	0.172	0.168	0.199	0.105	0.110	0.213	0.055	0.176	0.254	0.104	0.148	0.015	0.145
VBVR-Pro-FLUX.2	0.407	0.484	0.483	0.323	0.367	0.449	0.336	0.330	0.361	0.272	0.255	0.454	0.128
VBVR-Pro-Qwen-Image	0.322	0.332	0.298	0.217	0.193	0.431	0.222	0.311	0.341	0.239	0.233	0.413	0.181
Interleaved Image Generation Models
Proprietary Models
GPT-Image-2	0.507	0.428	0.456	0.318	0.428	0.206	0.300	0.587	0.398	0.413	0.633	0.480	0.303
Nano Banana Pro	0.564	0.480	0.518	0.422	0.512	0.285	0.174	0.648	0.553	0.499	0.657	0.585	0.220
Open-source Models
ThinkMorph-7B	0.154	0.113	0.100	0.082	0.101	0.148	0.031	0.195	0.176	0.166	0.163	0.253	0.103
VBVR-SenseNova-U1	0.408	0.469	0.356	0.313	0.373	0.386	0.477	0.347	0.291	0.317	0.275	0.480	0.238
SenseNova-U1-8B-MoT	0.565	0.533	0.501	0.395	0.544	0.355	0.349	0.597	0.448	0.495	0.533	0.717	0.401
Strong Baselines
VBVR-Pro-ThinkMorph	0.373	0.402	0.403	0.344	0.238	0.454	0.184	0.344	0.367	0.224	0.238	0.535	0.257
VBVR-Pro-SenseNova-U1	0.638	0.811	0.648	0.695	0.621	0.770	0.541	0.464	0.480	0.328	0.344	0.558	0.408
Video Generation Models
Proprietary Models
Veo 3.1	0.309	0.312	0.275	0.299	0.252	0.267	0.157	0.305	0.305	0.233	0.252	0.312	0.219
Kling V3	0.392	0.356	0.213	0.326	0.320	0.355	0.229	0.427	0.294	0.564	0.375	0.242	0.412
SeedDance 2.0	0.499	0.451	0.338	0.361	0.353	0.468	0.308	0.547	0.369	0.511	0.478	0.538	0.532
Open-source Models
HunyuanVideo-I2V	0.054	0.054	0.023	0.064	0.015	0.084	0.032	0.053	0.088	0.014	0.028	0.062	0.055
CogVideoX1.5-5B-I2V	0.085	0.100	0.061	0.118	0.069	0.092	0.060	0.070	0.125	0.038	0.051	0.040	0.024
Wan2.1-I2V-14B	0.100	0.105	0.052	0.125	0.091	0.102	0.052	0.095	0.112	0.073	0.071	0.123	0.044
Wan2.2-TI2V-5B	0.094	0.066	0.029	0.073	0.050	0.083	0.031	0.122	0.156	0.052	0.106	0.063	0.099
Wan2.2-I2V-14B-720P	0.182	0.157	0.082	0.131	0.110	0.161	0.156	0.207	0.224	0.139	0.140	0.195	0.273
LTX2.3-I2AV	0.112	0.106	0.062	0.109	0.070	0.133	0.055	0.119	0.161	0.135	0.086	0.091	0.050
VBVR-Wan2.2	0.517	0.548	0.237	0.499	0.334	0.566	0.591	0.486	0.310	0.343	0.345	0.732	0.684
Strong Baselines
VBVR-Pro-LTX2.3	0.425	0.527	0.409	0.510	0.346	0.460	0.390	0.324	0.381	0.108	0.201	0.477	0.386
VBVR-Pro-Wan2.1-I2V-14B	0.562	0.730	0.617	0.580	0.452	0.676	0.623	0.395	0.410	0.305	0.230	0.617	0.439
VBVR-Pro-Wan2.2-TI2V-5B	0.470	0.641	0.528	0.556	0.373	0.565	0.557	0.300	0.333	0.127	0.161	0.505	0.409
VBVR-Pro-Wan2.2-I2V-14B	0.670	0.808	0.632	0.685	0.556	0.751	0.636	0.532	0.479	0.418	0.350	0.679	0.690

## Quick Start ### Method 1: Standalone ThinkMorph inference #### Requirements - Linux and a CUDA-capable GPU - Python 3.10 - The [ThinkMorph](https://github.com/ThinkMorph/ThinkMorph) inference code and its dependencies For straightforward inference, an 80 GB-class GPU is recommended. The ThinkMorph loader can also distribute the model across multiple visible GPUs. ```bash git clone https://github.com/ThinkMorph/ThinkMorph.git cd ThinkMorph conda create -n thinkmorph python=3.10 -y conda activate thinkmorph pip install -r requirements.txt ``` #### Download ```python from huggingface_hub import snapshot_download snapshot_download( repo_id="Video-Reason/VBVR-Pro-ThinkMorph", local_dir="models/VBVR-Pro-ThinkMorph", allow_patterns=["*.json", "*.safetensors", "*.txt", "*.md"], ) ``` #### Inference Run the following from the ThinkMorph repository. The same downloaded directory is passed as both `model_path` and `config_path` because this release is self-contained. ```python import random import numpy as np import torch from PIL import Image from inferencer import InterleaveInferencer from scripts.cza.inference import load_model model_path = "models/VBVR-Pro-ThinkMorph" random.seed(42) np.random.seed(42) torch.manual_seed(42) torch.cuda.manual_seed_all(42) model, vae, tokenizer, token_ids, vae_transform, vit_transform = load_model( model_path, model_path, ) inferencer = InterleaveInferencer( model, vae, tokenizer, vae_transform, vit_transform, token_ids ) system_prompt = ( "Let's think step by step to answer the question. For text-based thinking, " "enclose the process within . For visual thinking, enclose " "the content within . Finally conclude with the " "final answer wrapped in " ) input_image = Image.open("first_frame.png").convert("RGB") output = inferencer( image=input_image, text="Move the object to the requested destination while preserving the scene.", understanding_output=False, think=True, system_prompt=system_prompt, max_think_token_n=4096, do_sample=True, text_temperature=0.3, cfg_text_scale=4.0, cfg_img_scale=2.0, cfg_interval=[0.0, 1.0], timestep_shift=3.0, num_timesteps=50, cfg_renorm_min=0.0, cfg_renorm_type="text_channel", max_rounds=10, ) for index, item in enumerate(output): if isinstance(item, Image.Image): item.save(f"generated_{index}.png") else: print(item, end="") ``` The settings above reproduce the first-attempt settings used by the existing VBVR-Pro evaluator: | Setting | Value | | --- | --- | | Seed | 42 | | Maximum thinking tokens | 4096 | | Text temperature | 0.3 | | Text CFG scale | 4.0 | | Image CFG scale | 2.0 | | Denoising steps | 50 | | Timestep shift | 3.0 | | Maximum interleaved rounds | 10 | ### Method 2: Unified VBVR-Pro inference Clone [Video-Reason/VBVR-Pro](https://github.com/Video-Reason/VBVR-Pro) and create its unified inference environment: ```bash git clone https://github.com/Video-Reason/VBVR-Pro.git cd VBVR-Pro/ uv sync --extra cu124 # or one of [cu118|cu121|cu124|cu126|cu128|cu129] source .venv/bin/activate ``` Then run the unified inference script: ```bash python example.py \ --model_path Video-Reason/VBVR-Pro-ThinkMorph \ --image_paths first_frame.png \ --prompt "Solve the task step by step." \ --think --max_rounds 10 --num_images 2 \ --output outputs ``` ## Notes - The model uses the custom ThinkMorph/BAGEL architecture and is not directly loadable with `transformers.AutoModel` or a Diffusers pipeline. - Inputs should follow the same image-plus-instruction format used by ThinkMorph. Generated images appear as `PIL.Image.Image` objects interleaved with text strings. - More generated images, larger resolutions, and longer reasoning traces increase runtime and memory use. - This checkpoint may inherit limitations and biases from its base model and fine-tuning data. Validate outputs for your application and use generated content responsibly. ## Citation ```bibtex @misc{gu2025thinkmorphemergentpropertiesmultimodal, title={ThinkMorph: Emergent Properties in Multimodal Interleaved Chain-of-Thought Reasoning}, author={Jiawei Gu and Yunzhuo Hao and Huichen Will Wang and Linjie Li and Michael Qizhe Shieh and Yejin Choi and Ranjay Krishna and Yu Cheng}, year={2025}, eprint={2510.27492}, archivePrefix={arXiv}, primaryClass={cs.CV}, url={https://arxiv.org/abs/2510.27492}, } ``` ## License This release follows the Apache 2.0 license declared by ThinkMorph-7B. See the base model repository for additional context.