SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2201.12096 · NeurIPS · 2022

Mask-based Latent Reconstruction for Reinforcement Learning

Tao Yu, Cuiling Lan, Zhizheng Zhang, Zhibo Chen, Yan Lu

arXiv · PDF · Open in the Atlas

Code that ran

8 of 9 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (4 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 6 have no contract verdict; 2 failures of our instrument; 1 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/2201.12096.json · badge: /api/ran/2201.12096.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
microsoft/Playvirtual community
Papers with Code archive
6 ran · 0 no recorded run
microsoft/Mask-based-Latent-Reconstruction marked official in Papers with Code
Papers with Code archive
2 ran · 1 no recorded run
FunctionStatusWhere it lives
Conv2dSame Ran
ran on a synthesized input · constructed an object; no result was computed
microsoft/Playvirtual/Atari/src/models.py
code served (permissive licence) · get_code("f7a03dfc44c34334")
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 an instance of its own class, Conv2dSame: the run constructed the object and did not call it, so no result was computed. 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

torch.manual_seed(0)

ARGS = {
    'in_channels': 3,
    'out_channels': 16,
    'kernel_size': 3,
    'bias': True,
    'stride': 1,
    'padding_layer': nn.ReflectionPad2d
}
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 Conv2dSame — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Conv2dSame 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
CubeMaskGenerator Ran
ran on a synthesized input
microsoft/Mask-based-Latent-Reconstruction/DMControl/src/mtm_sac.py
code served (permissive licence) · get_code("1ef174de7e17cfd1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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
import numpy as np

INIT_ARGS = {
    'input_size': (4, 6),
    'image_size': 24,
    'clip_size': 8,
    'block_size': 2,
    'mask_ratio': 0.5
}

ARGS = {}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float — no output shape is stated: observe_signature.py could not observe it (no_inputs)
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
PositionalEmbedding Ran
ran on a synthesized input
microsoft/Mask-based-Latent-Reconstruction/DMControl/src/mtm_sac.py
code served (permissive licence) · get_code("0b8b5b709ce22b1e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 3, dtype family float, shape (1, 16, 8) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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
import math

class PositionalEmbedding(nn.Module):

    def __init__(self, d_model, max_len=128):
        super().__init__()
        pe = torch.zeros(max_len, d_model).float()
        pe.require_grad = False
        position = torch.arange(0, max_len).float().unsqueeze(1)
        div_term = (torch.arange(0, d_model, 2).float() * -(math.log(10000.0) / d_model)).exp()
        pe[:, 0::2] = torch.sin(position * div_term)
        pe[:, 1::2] = torch.cos(position * div_term)
        pe = pe.unsqueeze(0)
        self.register_buffer('pe', pe)

    def forward(self, length):
        return self.pe[:, :length]


INIT_ARGS = {
    'd_model': 8,
    'max_len': 64
}

ARGS = {
    'length': 16
}

if __name__ == '__main__':
    model = PositionalEmbedding(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Output:\n{output}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (1, 16, 8) — shape (1, 16, 8) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
ResidualBlock Ran
ran on a synthesized input · constructed an object; no result was computed
microsoft/Playvirtual/Atari/src/models.py
code served (permissive licence) · get_code("bcd2e1b0c4219ca6")
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 an instance of its own class, ResidualBlock: the run constructed the object and did not call it, so no result was computed. 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

torch.manual_seed(0)

ARGS = {
    'in_channels': 3,
    'out_channels': 4,
    'norm_type': 'bn',
}
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 ResidualBlock — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ResidualBlock 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
RewardPredictor Ran
ran on a synthesized input · constructed an object; no result was computed
microsoft/Playvirtual/Atari/src/models.py
code served (permissive licence) · get_code("54f31a99fd9cb94b")
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 an instance of its own class, RewardPredictor: the run constructed the object and did not call it, so no result was computed. 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

torch.manual_seed(0)

# Define ARGS dict with all parameters for RewardPredictor
ARGS = {
    'input_channels': 3,
    'hidden_size': 8,
    'pixels': 36,
    'limit': 300,
    'norm_type': 'bn'
}
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 RewardPredictor — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a RewardPredictor 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
TransitionModel Ran
ran on a synthesized input · constructed an object; no result was computed
microsoft/Playvirtual/Atari/src/models.py
code served (permissive licence) · get_code("f9e622f31f261cbe")
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 an instance of its own class, TransitionModel: the run constructed the object and did not call it, so no result was computed. 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

torch.manual_seed(0)

ARGS = {
    'channels': 3,
    'num_actions': 4,
    'args': None,
    'blocks': 2,
    'hidden_size': 16,
    'pixels': 16,
    'limit': 10,
    'action_dim': 2,
    'norm_type': 'bn',
    'renormalize': True,
    'residual': False,
}
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 TransitionModel — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a TransitionModel 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
init_normalization Ran
ran; our contract draft was wrong, not the code
microsoft/Playvirtual/Atari/src/models.py
code served (permissive licence) · get_code("5298669e4fe00d40")
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 BatchNorm2d. 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

torch.manual_seed(0)

ARGS = {
    'channels': 32,
    'type': 'bn',
    'affine': True,
    'one_d': False
}
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 BatchNorm2d — a BatchNorm2d output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a BatchNorm2d 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
renormalize Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
microsoft/Playvirtual/Atari/src/models.py
code served (permissive licence) · get_code("fa63400bb2942968")
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 Tensor, ndim 3, dtype family float, shape (2, 4, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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

torch.manual_seed(0)

# Create a small tensor with shape (2, 4, 3)
# first_dim=1 means we'll normalize along dimension 1
tensor = torch.randn(2, 4, 3)

ARGS = {
    "tensor": tensor,
    "first_dim": 1
}
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 Tensor, ndim 3, dtype family float, shape (2, 4, 3) — shape (2, 4, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
MTM Not yet run
unverified (harvested, no recorded run)
microsoft/Mask-based-Latent-Reconstruction/DMControl/src/mtm_sac.py
code served (permissive licence) · get_code("7a1c73da025130f7")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 8 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
renormalize
brick of this paper's assembly: helper
3 other papers
3 later
2210.10763 (2022-10)
On the Feasibility of Cross-Task Transfer with Model-Based Reinforcement Learning
2404.05440 (2024-04)
Tree Search-Based Policy Optimization under Stochastic Execution Delay
2410.03119 (2024-10)
Spatial-Aware Decision-Making with Ring Attractors in Reinforcement Learning Systems

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

For deep reinforcement learning (RL) from pixels, learning effective state representations is crucial for achieving high performance. However, in practice, limited experience and high-dimensional inputs prevent effective representation learning. To address this, motivated by the success of mask-based modeling in other research fields, we introduce mask-based reconstruction to promote state representation learning in RL. Specifically, we propose a simple yet effective self-supervised method, Mask-based Latent Reconstruction (MLR), to predict complete state representations in the latent space from the observations with spatially and temporally masked pixels. MLR enables better use of context information when learning state representations to make them more informative, which facilitates the training of RL agents. Extensive experiments show that our MLR significantly improves the sample efficiency in RL and outperforms the state-of-the-art sample-efficient RL methods on multiple continuous and discrete control benchmarks. Our code is available at https://github.com/microsoft/Mask-based-Latent-Reconstruction.

For agents

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

get_harvested_code_for_paper("2201.12096")
get_code_for_paper("2201.12096")
have("2201.12096")

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

curl https://syntology.ai/api/ran/2201.12096.json

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

[![Syntology run record](https://syntology.ai/api/ran/2201.12096.svg)](https://syntology.ai/paper/2201.12096)

Connect an agent — have() is free.