SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.15922 · 2026

World Action Models are Zero-shot Policies

arXiv · PDF · Open in the Atlas

Code that ran

2 of 6 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 failures of our instrument; 4 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2602.15922.json · badge: /api/ran/2602.15922.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
dreamzero0/dreamzero canonical (classifier label)
classifier basis: census
2 ran · 4 no recorded run
FunctionStatusWhere it lives
ensure_file Ran
ran; our contract draft was wrong, not the code
dreamzero0/dreamzero/groot/vla/model/dreamzero/action_head/wan_flow_matching_action_tf.py
code served (permissive licence) · get_code("c7fc5bf1b143ecf1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import tempfile
from unittest.mock import patch, MagicMock

# Create a temporary file to use as a valid local path
temp_dir = tempfile.gettempdir()
temp_file = os.path.join(temp_dir, "test_model.pt")

# Write a dummy file so os.path.exists returns True
with open(temp_file, 'w') as f:
    f.write("dummy model")

# Mock the hf_hub_download to avoid actual network calls
mock_hf_download = MagicMock(return_value="/mock/cache/path/model.pt")

ARGS = {
    "path": temp_file,
    "hf_filename": "model.safetensors",
    "repo_id": "Wan-AI/Wan2.1-I2V-14B-480P"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
huggingface_hub — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
restore_from_saved Ran
ran; our contract draft was wrong, not the code
dreamzero0/dreamzero/groot/vla/model/dreamzero/modules/cudnn_attention.py
code served (permissive licence) · get_code("85c20bab86009e1f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from typing import Any, List, Optional, Tuple, Union

torch.manual_seed(0)

# Create a mock object that implements restore_from_saved method
class MockTensorWithRestore:
    def __init__(self, value: int):
        self.value = value
    
    def restore_from_saved(self, saved_tensors):
        # This mock consumes one tensor from saved_tensors and returns the rest
        return saved_tensors[1:]

# Create test data
test_tensor_1 = torch.randn(2, 4)
test_tensor_2 = torch.randn(2, 4)
test_tensor_3 = torch.randn(2, 4)
test_tensor_4 = torch.randn(2, 4)

# Create the saved_tensors tuple - these will be consumed during restoration
saved_tensors = (test_tensor_1, test_tensor_2, test_tensor_3, test_tensor_4)

# Create tensors list with mix of None, torch.Tensor, and custom objects
tensors = [
    None,                           # Will consume one saved_tensor
    test_tensor_2,                  # Will consume one saved_tensor (though it's already a tensor)
    MockTensorWithRestore(42),      # Will call restore_from_saved, consuming one saved_tensor
    None                            # Will consume one saved_tensor
]

ARGS = {
    "tensors": tensors,
    "saved_tensors": saved_tensors
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
attention Not yet run
unverified (harvested, no recorded run)
dreamzero0/dreamzero/groot/vla/model/dreamzero/modules/attention.py
code served (permissive licence) · get_code("6f8f364555e907ac")
How it ran

no invocation recorded.

create_vla_with_pretrained_action_head Not yet run
unverified (harvested, no recorded run)
dreamzero0/dreamzero/groot/vla/model/dreamzero/base_vla.py
code served (permissive licence) · get_code("379ec7eb7a93d7a5")
How it ran

no invocation recorded.

flash_attention Not yet run
unverified (harvested, no recorded run)
dreamzero0/dreamzero/groot/vla/model/dreamzero/modules/attention.py
code served (permissive licence) · get_code("5f985627e201aeeb")
How it ran

no invocation recorded.

hf_download Not yet run
unverified (harvested, no recorded run)
dreamzero0/dreamzero/groot/vla/model/dreamzero/action_head/wan_flow_matching_action_tf.py
code served (permissive licence) · get_code("838185d97644cee9")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 1 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 0 later) and 5 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
attention 3 other papers
3 dated earlier
2502.11079 (2025-02)
Phantom: Subject-Consistent Video Generation via Cross-Modal Alignment
2505.23742 (2025-05)
MAGREF: Masked Guidance for Any-Reference Video Generation with Subject Disentanglement
2507.06119 (2025-07)
Omni-Video: Democratizing Unified Video Understanding and Generation

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

-Language-Action (VLA) models excel at semantic generalization but struggle to generalize to unseen physical motions in novel environments. We introduce DreamZero, a World Action Model (WAM) built upon a pretrained video diffusion backbone. Unlike VLAs, WAMs learn physical dynamics by predicting future world states and actions, using video as a dense representation of how the world evolves. By jointly modeling video and action, DreamZero learns diverse skills effectively from heterogeneous robot data without relying on repetitive demonstrations. This results in over 2× improvement in generalization to new tasks and environments compared to state-of-the-art VLAs in realrobot experiments. Crucially, through model and system optimizations, we enable a 14B autoregressive video diffusion model to perform real-time closed-loop control at 7Hz. Finally, we demonstrate two forms of cross-embodiment transfer: video-only demonstrations from other robots or humans yield a relative improvement of over 42% on unseen task performance with just 10-20 minutes of data. More surprisingly, DreamZero enables few-shot embodiment adaptation, transferring to a new embodiment with only 30 minutes of play data while retaining zero-shot generalization.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2602.15922")
get_code_for_paper("2602.15922")
have("2602.15922")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2602.15922.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2602.15922.svg)](https://syntology.ai/paper/2602.15922)

Connect an agent — have() is free.