SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2212.06331 · 2022

DeepMapping2: Self-Supervised Large-Scale LiDAR Map Optimization

arXiv · PDF · Open in the Atlas

Code that ran

9 of 9 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 9 have no contract verdict; 0 failures of our instrument; 0 no recorded run), measured 2026-11-21; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; 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/2212.06331.json · badge: /api/ran/2212.06331.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
ai4ce/DeepMapping2 canonical (classifier label)
classifier basis not recorded on the link
9 ran · 0 no recorded run
FunctionStatusWhere it lives
bce Ran
ran on a synthesized input
ai4ce/DeepMapping2/loss/bce_loss.py
code served (permissive licence) · get_code("cf37e143cc15ae8d")
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 0, dtype family float, shape () 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.functional as F

torch.manual_seed(0)

# Create small tensors for binary cross entropy with logits
# pred: logits (unbounded), shape [batch_size, num_classes]
# targets: binary labels (0 or 1), same shape as pred
batch_size = 2
num_classes = 4

pred = torch.randn(batch_size, num_classes)
targets = torch.randint(0, 2, (batch_size, num_classes)).float()

# weight: optional per-sample or per-class weights
# Can be None or a tensor of shape matching pred/targets or broadcastable
weight = None

ARGS = {
    "pred": pred,
    "targets": targets,
    "weight": weight,
}
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 0, dtype family float, shape () — shape () 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
bce_with_logits Ran
ran on a synthesized input
ai4ce/DeepMapping2/loss/bce_loss.py
code served (permissive licence) · get_code("c180601a5f83d9c4")
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 0, dtype family float, shape () 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

torch.manual_seed(0)

# Construct tiny tensors for the function
batch_size = 2
dim = 4

# input: logits tensor
input_tensor = torch.randn(batch_size, dim)

# target: binary labels (same size as input)
target_tensor = torch.randint(0, 2, (batch_size, dim)).float()

# weight: optional weight tensor (same size as input)
weight_tensor = torch.ones(batch_size, dim)

ARGS = {
    'input': input_tensor,
    'target': target_tensor,
    'weight': weight_tensor,
    'reduction': 'elementwise_mean'
}
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 0, dtype family float, shape () — shape () 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
chamfer_loss Ran
ran on a synthesized input
ai4ce/DeepMapping2/loss/chamfer_dist.py
code served (permissive licence) · get_code("fb4d5906d5f4e5fb")
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 0, dtype family float, shape () 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

torch.manual_seed(0)

# Create sample tensors
# obs shape: <BxLx2> where B is batch, L is sequence length, 2 is spatial dims
# For seq=2, we need at least 2 frames, so let's use B=3, L=4
B = 3
L = 4
spatial_dim = 2

obs = torch.randn(B, L, spatial_dim)

# valid_obs shape: <BxL> with binary validity indicators
valid_obs = torch.ones(B, L, dtype=torch.long)
# Mark some points as invalid for testing
valid_obs[0, 1] = 0
valid_obs[1, 3] = 0

# seq parameter - number of consecutive frames to consider
seq = 2

ARGS = {
    "obs": obs,
    "valid_obs": valid_obs,
    "seq": seq,
}
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 0, dtype family float, shape () — shape () 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
find_valid_points Ran
ran on a synthesized input
ai4ce/DeepMapping2/dataset_loader/KITTI.py
code served (permissive licence) · get_code("7cf01cf5dee27e00")
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 2, dtype family float, shape (2, 4) 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 valid point cloud tensor with shape (B, N, k)
# B=2 (batch size), N=4 (number of points), k=3 (coordinates)
local_point_cloud = torch.tensor([
    [[1.0, 2.0, 3.0],
     [0.0, 0.0, 0.0],
     [0.5, 0.0, 0.0],
     [0.0, 0.0, 0.0]],
    [[0.0, 0.0, 0.0],
     [1.5, 2.5, 3.5],
     [0.0, 0.0, 0.0],
     [0.1, 0.2, 0.3]]
])

ARGS = {
    "local_point_cloud": local_point_cloud
}
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 2, dtype family float, shape (2, 4) — shape (2, 4) 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
get_MLP_layers Ran
ran on a synthesized input
ai4ce/DeepMapping2/models/networks.py
code served (permissive licence) · get_code("074261dae92b8e6c")
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 list, ndim 1, 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

torch.manual_seed(0)

# The function get_MLP_layers takes:
# - dims: a list/sequence of integers representing layer dimensions
# - doLastRelu: a boolean indicating whether to apply ReLU after the last layer

ARGS = {
    "dims": [4, 8, 6, 3],
    "doLastRelu": 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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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
get_M_net_inputs_labels Ran
ran on a synthesized input
ai4ce/DeepMapping2/models/deepmapping.py
code served (permissive licence) · get_code("4add307b57676d65")
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

torch.manual_seed(0)

# Create small tensors for occupied and unoccupied points
# batch_size=2, n_occupied=3, n_unoccupied=4, coordinates=5
batch_size = 2
n_occupied = 3
n_unoccupied = 4
coord_dim = 5

occupied_points = torch.randn(batch_size, n_occupied, coord_dim)
unoccupited_points = torch.randn(batch_size, n_unoccupied, coord_dim)

ARGS = {
    "occupied_points": occupied_points,
    "unoccupited_points": unoccupited_points,
}
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
get_and_init_FC_layer Ran
ran on a synthesized input
ai4ce/DeepMapping2/models/networks.py
code served (permissive licence) · get_code("741a7fb5e7961d1a")
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 Linear. 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)

# The function takes two integer parameters: din (input dimension) and dout (output dimension)
# It creates a Linear layer, initializes weights with Xavier uniform, and fills bias with 0
# We need to provide small but distinct integer dimensions

din = 4
dout = 3

ARGS = {
    'din': din,
    'dout': dout
}
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 Linear — a Linear output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Linear 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
registration_loss Ran
ran on a synthesized input
ai4ce/DeepMapping2/loss/chamfer_dist.py
code served (permissive licence) · get_code("14f4db8371a0b16f")
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 0, dtype family float, shape () 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

torch.manual_seed(0)

# Create obs tensor: shape (B, L, 2) where B=batch, L=sequence length, 2=coordinates
# Using B=2, L=4, K=2 (last dimension)
B = 2
L = 4
K = 2

obs = torch.randn(B, L, K, dtype=torch.float32)

# Create valid_obs tensor: shape (B, L) with binary validity masks
valid_obs = torch.randint(0, 2, (B, L), dtype=torch.int64)
# Ensure at least one valid point per batch for division safety
valid_obs[0, 0] = 1
valid_obs[1, 0] = 1

ARGS = {
    "obs": obs,
    "valid_obs": valid_obs,
}
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 0, dtype family float, shape () — shape () 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
sample_unoccupied_point Ran
ran on a synthesized input
ai4ce/DeepMapping2/models/deepmapping.py
code served (permissive licence) · get_code("e82696f20df7f8af")
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, 20, 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)

# Parameters:
# local_point_cloud: tensor of shape <BxLxk> representing point cloud
# n_samples: integer number of samples on each ray

local_point_cloud = torch.randn(2, 4, 3)  # B=2, L=4, k=3
n_samples = 5

ARGS = {
    "local_point_cloud": local_point_cloud,
    "n_samples": n_samples,
}
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, 20, 3) — shape (2, 20, 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

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, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 9 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.

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 2026-11-21.

Repositories linked to this paper

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

Abstract

LiDAR mapping is important yet challenging in self-driving and mobile robotics. To tackle such a global point cloud registration problem, DeepMapping converts the complex map estimation into a self-supervised training of simple deep networks. Despite its broad convergence range on small datasets, DeepMapping still cannot produce satisfactory results on large-scale datasets with thousands of frames. This is due to the lack of loop closures and exact cross-frame point correspondences, and the slow convergence of its global localization network. We propose DeepMapping2 by adding two novel techniques to address these issues: (1) organization of training batch based on map topology from loop closing, and (2) self-supervised local-to-global point consistency loss leveraging pairwise registration. Our experiments and ablation studies on public datasets (KITTI, NCLT, and Nebula) demonstrate the effectiveness of our method.

For agents

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

get_harvested_code_for_paper("2212.06331")
get_code_for_paper("2212.06331")
have("2212.06331")

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

curl https://syntology.ai/api/ran/2212.06331.json

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

[![Syntology run record](https://syntology.ai/api/ran/2212.06331.svg)](https://syntology.ai/paper/2212.06331)

Connect an agent — have() is free.