Belief4D-WM-checkpoints

polikwt/Belief4D-WM-checkpoints

Belief4D-WM-checkpoints is a machine learning model from polikwt. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for pytorch.

This repository stores checkpoints for https://github.com/Destroyertubu/Belief4D-WM at source commit f8e4be4d4f35f7dc04ba9b235810f4e25ce50516.

Downloads · 30 days

Access

Public

Updated Jul 21, 2026

Repo size

4.5 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.pt4.5 GB · 100%

At a glance

Library: pytorch
Access: Public
Created: Jul 15, 2026
Updated: Jul 21, 2026
SHA: 0daf3e3b

Model card

From the Hugging Face model README

Belief4D-WM Checkpoint Registry

This repository stores checkpoints for https://github.com/Destroyertubu/Belief4D-WM at source commit f8e4be4d4f35f7dc04ba9b235810f4e25ce50516.

Every checkpoint is a pilot, development, or privileged-component artifact. The paper primary claim remains NOT_EVALUATED; these files must not be presented as end-to-end manuscript results.

Safety

The files are PyTorch training checkpoints and can contain pickle payloads. Load them only from this trusted repository after verifying SHA256. Checkpoints may include model, optimizer, configuration, history, and RNG state.

Files

Run	Phase	Size	Status	SHA256
`gpu-20260714-063459`	`t1-t2-privileged-state`	62.72 MiB	`archived_pilot_non_claim`	`f64bd20ad89fd87b458a911f4b899e7aaf475936af756819698ca00624515d46`
`gpu-20260714-063625`	`t1-t2-privileged-state`	62.72 MiB	`archived_pilot_non_claim`	`b30304a5009d136105dd40a601de943790656025339c4458f60b8de746cdf23f`
`gpu-20260714-063759`	`t1-t2-privileged-state`	62.73 MiB	`archived_pilot_non_claim`	`4a2ac5c56001cda8d69e535fdf09023863fa555f02b53776c3934c8ab85067da`
`gpu-20260714-064246`	`t1-t2-privileged-state`	62.74 MiB	`archived_pilot_non_claim`	`af118c655b19f214eed527cf785f0464a9a4476a83c229b00c4f3e98bf015eda`
`t3-20260714-102305`	`t3-learned-evidence`	13.43 MiB	`archived_pilot_non_claim`	`721af1698c16cd9dc551446666817b7674f9bb9d7f2c20c29170c5a52ef98672`
`t3-20260714-103214`	`t3-learned-evidence`	13.43 MiB	`archived_pilot_non_claim`	`e58f603260e0001b8a8c9a8add965f0ca691c04cd8852f0dcc3194e4bb0d49be`
`t3-20260714-114612`	`t3-learned-evidence`	13.45 MiB	`archived_pilot_non_claim`	`59f682aa9089d3e715c6f0f00ff8b84947ca1b181f4f1c3451f5f7071efe9345`
`t3-20260714-115327`	`t3-learned-evidence`	13.45 MiB	`archived_pilot_non_claim`	`18661933e043852d1371de042cdee84dbf0a1525c963fa158d0c692c05f1cd5b`
`t6-20260716-062015`	`pilot/t6-geometry-renderer`	6.14 MiB	`verified_pilot_non_claim`	`f58d3b9bf8caba7976b6a8b31ef3e6b677eddcf1b3a981cb0b2d430c111b3f89`
`t7d-20260717-060014`	`kubric/development-learned-evidence`	74.74 MiB	`audited_development_non_claim`	`ed6320dfaaecfcf7299961d4858fdeef7194b644d1f265ce5811f396768b3432`
`t7d-20260719-111319`	`kubric/development-learned-evidence`	74.74 MiB	`audited_development_non_claim`	`b4ec403f8a2643cd2629224f867e809bc0fac42da8347ad2224aa796f5e20292`
`t2k-20260719-121203`	`kubric/development-privileged-dynamics`	111.22 MiB	`audited_development_non_claim`	`a63dc6034936c99b0cc331f46ce15de0263c12ccfa3a74c30a6f4a7f20c209bc`
`t2k-20260719-134943`	`kubric/development-privileged-dynamics`	111.22 MiB	`audited_development_non_claim`	`492a04259521ab5193aac01c6a8dabad4c29bacd9b5680b12f923bf38ac98407`
`t2k-20260719-135323`	`kubric/development-privileged-dynamics`	111.22 MiB	`audited_development_non_claim`	`8f18916cfb4a95acdc715ae4674001376dcab5fc5cd817460bc62d8a66aaedce`
`t2k-20260719-135728`	`kubric/development-privileged-dynamics`	111.22 MiB	`audited_development_non_claim`	`bfdd11c963f54ec33a1962ae5921a6a7972cff7f693e1ddeb63f0686650b62a5`
`t2k-20260719-140123`	`kubric/development-privileged-dynamics`	111.22 MiB	`audited_development_non_claim`	`d8984a3ed15b07291b86209a8e8499f39fc58d1bdb3500189aac0f88536451aa`
`pfit-20260720-020446`	`kubric/locked-privileged-component`	3280.30 MiB	`locked_component_non_claim`	`25d7e63bf183ef2c90b3ea3156055369d9026d69b7bc3308802b018cd9b9799a`

Machine-readable metadata is in checkpoint_index.json; CHECKSUMS.sha256 verifies repository-relative checkpoint paths.

Scope

The registry intentionally keeps successful smoke, audited development, and locked privileged-component checkpoints in separate phase directories. Reports linked from the index define inputs, metrics, limitations, and audit status. Quick probes and failed attempts remain in the GitHub run ledger but are not promoted to this model registry.

No license is asserted by this model repository; consult the source repository owner before redistribution or derivative release.

Library: pytorch
Created: Jul 15, 2026
Updated: Jul 21, 2026