World4Scorer by Jieyuan: Open-Weight Model

Model Card

By Jieyuan, published under apache-2.0, revision 593ead343060.

Jieyuan Pei, Meiyi Lu, Sining Ang, Yubo Zhao, Zhangyi Hu, Mingwei Xu, Haokai Ding, Wei Li, Zihan You, Jianwei Zheng, Li Yu, Yifeng Pan, Ji Tao, Rongjunchen Zhang, Yan Wang Checkpoints of World4Scorer (paper, code). World4Scorer makes the candidate scorer of a generate-and-select planner a trajectory-conditioned latent world model. One shared predictor gives every candidate trajectory a predicted state, and score heads read its outcomes (collision, drivable area, progress, time to collision, comfort) from that state. Simulator outcomes supervise every candidate; the one future the driving log actually recorded anchors the shared predictor during training. No future is needed at test time.…

Read Jieyuan's full model card

### Outcome-Grounded World Modeling for Autonomous Driving [Project page](https://world4scorer.github.io/) · [Code (GitHub)](https://github.com/World4Scorer/World4Scorer) · [Film](https://world4scorer.github.io/#video) · [Paper (arXiv)](https://arxiv.org/abs/2609.36438) Jieyuan Pei, Meiyi Lu, Sining Ang, Yubo Zhao, Zhangyi Hu, Mingwei Xu, Haokai Ding, Wei Li, Zihan You, Jianwei Zheng, Li Yu, Yifeng Pan, Ji Tao, Rongjunchen Zhang, Yan Wang

_{The 95-second overview film, with sound. Also on the project page.}

About

Checkpoints of World4Scorer (paper, code).

World4Scorer makes the candidate scorer of a generate-and-select planner a trajectory-conditioned latent world model. One shared predictor gives every candidate trajectory a predicted state, and score heads read its outcomes (collision, drivable area, progress, time to collision, comfort) from that state. Simulator outcomes supervise every candidate; the one future the driving log actually recorded anchors the shared predictor during training. No future is needed at test time.

Results

Benchmark	Metric	World4Scorer	Best other method in the paper
NAVSIM-v2 navtest	EPDMS	93.0 (91.4 without inertial re-ranking)	Drive-JEPA 90.8
NAVSIM-v1 navtest	PDMS	94.0	DrivoR 93.7 · Drive-JEPA 93.7
Bench2Drive-220, closed loop	Driving Score	73.27	ReCogDrive 71.36
OGBench-Cube, LeWM planner	Success rate	73.64%	LeWM 68.91% (same world model and budget)

The Bench2Drive system is adapted (official 1000-clip training subset, a route point and two deployment rules). Cube numbers are our runs over 11 seeds × 50 episodes.

Demos

_{NAVSIM, Las Vegas. 64 candidates every 0.5 s, coloured by predicted score.}	_{OGBench-Cube. Left: LeWM alone fails. Right: with the outcome score it succeeds.}
_{Bench2Drive. A taxi cuts into the lane (closed loop, 1.5×).}	_{Bench2Drive. A pedestrian steps into the road at night (closed loop, 1.5×).}

Checkpoints

File	Model	Result	SHA-256
`weights/w4s_navsim_ep23.ckpt`	NAVSIM planner, seed 3, epoch 23, 35.5M parameters	91.4 EPDMS · 93.0 with re-ranking · 94.0 PDMS	`22604f83dda8a2bd2dd05356fca8457f0c7dc78439f47db396e5c85d74971a44`
`weights/w4s_b2d_route_tp_ep14.ckpt`	Bench2Drive planner, route point, epoch 14	73.27 Driving Score	`ed3f61a9ac850627c07bbf1d37cc0a9e123d5b41a9c1cecefe820644e14aff5b`
`weights/w4s_b2d_route_blind_ep14.ckpt`	Bench2Drive route-blind model, epoch 14	initialization of the route-point model	`640dd42934b57d9305c336db7234747533977ad916c18723a9bb97b385407d31`
`ogbench_cube/heads/n10_r1/head.pt`	OGBench-Cube outcome head `n10_r1`	73.64% success	`fce54aa53576587f002a8a515db393bc346b1daee2716630ecf14352fb08bf0e`
`weights/realized_future_emb_navtrain_f7_camf0.pt`	NAVSIM future targets (navtrain)	used only in training	`2d64716d8f07a0dbc94e3f240108167acf9c9ef5093635eacb857725a13f0d82`

From the root of the code repository, this puts every file where the scripts expect it:

hf download pei2333/World4Scorer --include "weights/*" --include "ogbench_cube/*" --local-dir .

The SHA-256 values identify the exact files behind the reported numbers; check a download with python tools/verify_release.py --artifact KIND=PATH in the code repository.

Citation

@article{pei2026world4scorer,
  title   = {World4Scorer: Outcome-Grounded World Modeling for Autonomous Driving},
  author  = {Pei, Jieyuan and Lu, Meiyi and Ang, Sining and Zhao, Yubo and Hu, Zhangyi and Xu, Mingwei and Ding, Haokai and Li, Wei and You, Zihan and Zheng, Jianwei and Yu, Li and Pan, Yifeng and Tao, Ji and Zhang, Rongjunchen and Wang, Yan},
  journal = {arXiv preprint arXiv:2609.36438},
  year    = {2026}
}

Identity and Version

Repository: pei2333/World4Scorer
Publisher: Jieyuan
Task: Not stated by the source
Modality: Other
Library: Not stated by the source
Parameters: Not stated by the source
Languages: en
Revision: 593ead3430609ecd532bc486986bdc9557d8a082
First published: 2026-09-27
Last updated: 2026-09-30

Files and Weights

15 files, 655.7 MB in total. The weights are 5 files totalling 640.7 MB in ckpt, pt.

Weights5 files · 640.7 MB

Documentation1 file · 5.0 KB

Other8 files · 15.0 MB

Repository1 file · 1.9 KB

Every file

File	Type	Size	SHA-256
ogbench_cube/heads/n10_r1/head.pt	Weights	983.1 KB	fce54aa53576
weights/realized_future_emb_navtrain_f7_camf0.pt	Weights	112.8 MB	2d64716d8f07
weights/w4s_b2d_route_blind_ep14.ckpt	Weights	142.3 MB	640dd42934b5
weights/w4s_b2d_route_tp_ep14.ckpt	Weights	142.3 MB	ed3f61a9ac85
weights/w4s_navsim_ep23.ckpt	Weights	242.3 MB	22604f83dda8
README.md	Documentation	5.0 KB	—
assets/demo_cube.webp	Other	438.2 KB	7cda1b4ca1a3
assets/demo_cutin.webp	Other	1.0 MB	e6c78e901116
assets/demo_navsim.webp	Other	1.3 MB	1452bf55c0a3
assets/demo_night.webp	Other	1.4 MB	7003794682cf
assets/film.mp4	Other	9.9 MB	0e478a6b4d90
assets/hero.webp	Other	765.6 KB	037521079b81
assets/method.webp	Other	145.9 KB	f64602cac3fc
assets/wordmark.svg	Other	8.3 KB	—
.gitattributes	Repository	1.9 KB	—

License and Download

License: apache-2.0
Access: Open weights, no gate
Download size: 640.7 MB

Download from Jieyuan

Released by Jieyuan through its official repository on Hugging Face. Read the license.

Built From

Described by arXiv:2609.36438

Memory Requirements

Precision	Weights in memory
As published	640.7 MB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About World4Scorer

Can I use World4Scorer commercially?

Yes. World4Scorer is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.