SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2111.06389 · 2021

Full-Body Visual Self-Modeling of Robot Morphologies

arXiv · PDF · Open in the Atlas

Code that ran

6 of 8 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; 1 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 1 failure of our instrument; 2 no recorded run), measured 2026-11-22; 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/2111.06389.json · badge: /api/ran/2111.06389.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
BoyuanChen/visual-selfmodeling canonical (classifier label)
classifier basis not recorded on the link
6 ran · 2 no recorded run
FunctionStatusWhere it lives
compute_unit_sphere_transform Ran
ran on a synthesized input
BoyuanChen/visual-selfmodeling/utils/common.py
code served (permissive licence) · get_code("2a0c158381f5db93")
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 numpy as np

class BoundingBox:
    def __init__(self, centroid):
        self.centroid = centroid

class Mesh:
    def __init__(self, vertices, centroid):
        self.vertices = vertices
        self.bounding_box = BoundingBox(centroid)

# Create a simple mesh with 4 vertices in 3D space
vertices = np.array([
    [1.0, 0.0, 0.0],
    [-1.0, 0.0, 0.0],
    [0.0, 1.0, 0.0],
    [0.0, -1.0, 0.0]
], dtype=np.float32)

centroid = np.array([0.0, 0.0, 0.0], dtype=np.float32)

mesh = Mesh(vertices, centroid)

ARGS = {
    "mesh": mesh
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — 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_body_colors Ran
ran on a synthesized input
BoyuanChen/visual-selfmodeling/utils/common.py
code served (permissive licence) · get_code("f3a2b8fb9bd02cc7")
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 dict. 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 numpy as np

# Create a mock client class that simulates PyBullet's behavior
class MockPyBulletClient:
    def getVisualShapeData(self, body_id):
        # Returns a list of visual shape data tuples
        # Each tuple contains visual info where:
        # - index 1 is the link index
        # - index 7 is the color (RGBA tuple)
        return [
            (0, 0, 0, 0, 0, 0, 0, (1.0, 0.0, 0.0, 1.0), 0),  # link 0, red color
            (1, 1, 0, 0, 0, 0, 0, (0.0, 1.0, 0.0, 1.0), 0),  # link 1, green color
            (2, 2, 0, 0, 0, 0, 0, (0.0, 0.0, 1.0, 1.0), 0),  # link 2, blue color
        ]

client = MockPyBulletClient()
body_id = 0

ARGS = {
    "client": client,
    "body_id": body_id
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
get_view_frustum Ran
ran, honoured the contract we drafted
BoyuanChen/visual-selfmodeling/fusion.py
code served (permissive licence) · get_code("4c74437ab5ad2ac8")
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 ndarray, ndim 2, dtype family float, shape (3, 5) 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 numpy as np

# Create a small depth image (height=4, width=5)
depth_im = np.random.rand(4, 5).astype(np.float32) * 5.0 + 1.0

# Create camera intrinsics matrix (3x3)
# cam_intr[0, 0] = fx, cam_intr[0, 2] = cx
# cam_intr[1, 1] = fy, cam_intr[1, 2] = cy
cam_intr = np.array([
    [500.0, 0.0, 2.5],
    [0.0, 500.0, 2.0],
    [0.0, 0.0, 1.0]
], dtype=np.float32)

# Create camera pose (4x4 homogeneous transformation matrix)
cam_pose = np.array([
    [1.0, 0.0, 0.0, 0.1],
    [0.0, 1.0, 0.0, 0.2],
    [0.0, 0.0, 1.0, 0.3],
    [0.0, 0.0, 0.0, 1.0]
], dtype=np.float32)

ARGS = {
    "depth_im": depth_im,
    "cam_intr": cam_intr,
    "cam_pose": cam_pose
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (3, 5) — shape (3, 5) 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
load_config Ran
ran on a synthesized input
BoyuanChen/visual-selfmodeling/utils/eval_mesh.py
code served (permissive licence) · get_code("cbbf6e45e2f6ff75")
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 dict. 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 tempfile
import os
import yaml

# Create a temporary YAML file with valid trainer parameters
temp_dir = tempfile.mkdtemp()
yaml_filepath = os.path.join(temp_dir, "config.yaml")

config_data = {
    "learning_rate": 0.001,
    "batch_size": 32,
    "epochs": 10,
    "model": {
        "type": "transformer",
        "hidden_dim": 256,
        "num_layers": 2
    },
    "optimizer": "adam",
    "loss_fn": "cross_entropy"
}

with open(yaml_filepath, 'w') as f:
    yaml.dump(config_data, f)

ARGS = {
    "filepath": yaml_filepath
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
yaml — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
rename_ckpt_for_multi_models Ran
ran on a synthesized input
BoyuanChen/visual-selfmodeling/models.py
code served (permissive licence) · get_code("faf47cf1a0889190")
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 OrderedDict. 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 collections import OrderedDict

torch.manual_seed(0)

# Create a minimal checkpoint dict with state_dict containing model keys
ckpt = {
    'state_dict': OrderedDict([
        ('model.layer1.weight', torch.randn(4, 3)),
        ('model.layer1.bias', torch.randn(4)),
        ('model.layer2.weight', torch.randn(5, 4)),
        ('other_key', torch.randn(2, 2))
    ])
}

ARGS = {
    'ckpt': ckpt
}
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 OrderedDict — a OrderedDict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a OrderedDict 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
rigid_transform Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
BoyuanChen/visual-selfmodeling/fusion.py
code served (permissive licence) · get_code("a1150c1b03c31b42")
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 ndarray, ndim 2, dtype family float, shape (2, 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 numpy as np

# Create a small pointcloud with N=2 points, each with (x, y, z) coordinates
xyz = np.array([
    [1.0, 2.0, 3.0],
    [4.0, 5.0, 6.0]
], dtype=np.float32)

# Create a 4x4 homogeneous transformation matrix (rotation + translation)
# This is a simple transformation: identity rotation with translation (0.1, 0.2, 0.3)
transform = np.array([
    [1.0, 0.0, 0.0, 0.1],
    [0.0, 1.0, 0.0, 0.2],
    [0.0, 0.0, 1.0, 0.3],
    [0.0, 0.0, 0.0, 1.0]
], dtype=np.float32)

ARGS = {
    "xyz": xyz,
    "transform": transform
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (2, 3) — shape (2, 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
check_mesh_contains Not yet run
unverified (harvested, no recorded run)
BoyuanChen/visual-selfmodeling/utils/libmesh/inside_mesh.py
code served (permissive licence) · get_code("3c5c35d8c716ebad")
How it ran

no invocation recorded.

gradient Not yet run
unverified (harvested, no recorded run)
BoyuanChen/visual-selfmodeling/utils/common.py
code served (permissive licence) · get_code("72e1c034735c23ca")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 6 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 5 later) and 2 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
rigid_transform 9 other papers
9 later
2112.00236 (2021-12)
VoRTX: Volumetric 3D Reconstruction With Transformers for Voxelwise View Selection and Fusion
2204.01139 (2022-04)
BNV-Fusion: Dense 3D Reconstruction using Bi-level Neural Volume Fusion
2207.11514 (2022-07)
Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Models
and 6 more not named here
get_view_frustum 8 other papers
8 later
2112.00236 (2021-12)
VoRTX: Volumetric 3D Reconstruction With Transformers for Voxelwise View Selection and Fusion
2204.01139 (2022-04)
BNV-Fusion: Dense 3D Reconstruction using Bi-level Neural Volume Fusion
2207.11514 (2022-07)
Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Models
and 5 more not named here
gradient 2 other papers
2 later
2407.09679 (2024-07)
Physics-Informed Learning of Characteristic Trajectories for Smoke Reconstruction
2409.19472 (2024-09)
Towards Croppable Implicit Neural Representations
check_mesh_contains 1 other paper
1 dated earlier
2105.02788 (2021-05)
ACORN: Adaptive Coordinate Networks for Neural Scene Representation
load_config 1 other paper
1 later
2112.10755 (2021-12)
Discovering State Variables Hidden in Experimental Data
rename_ckpt_for_multi_models 1 other paper
1 later
2112.10755 (2021-12)
Discovering State Variables Hidden in Experimental Data

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

Internal computational models of physical bodies are fundamental to the ability of robots and animals alike to plan and control their actions. These "self-models" allow robots to consider outcomes of multiple possible future actions, without trying them out in physical reality. Recent progress in fully data-driven self-modeling has enabled machines to learn their own forward kinematics directly from task-agnostic interaction data. However, forward-kinema\-tics models can only predict limited aspects of the morphology, such as the position of end effectors or velocity of joints and masses. A key challenge is to model the entire morphology and kinematics, without prior knowledge of what aspects of the morphology will be relevant to future tasks. Here, we propose that instead of directly modeling forward-kinematics, a more useful form of self-modeling is one that could answer space occupancy queries, conditioned on the robot's state. Such query-driven self models are continuous in the spatial domain, memory efficient, fully differentiable and kinematic aware. In physical experiments, we demonstrate how a visual self-model is accurate to about one percent of the workspace, enabling the robot to perform various motion planning and control tasks. Visual self-modeling can also allow the robot to detect, localize and recover from real-world damage, leading to improved machine resiliency. Our project website is at: https://robot-morphology.cs.columbia.edu/

For agents

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

get_harvested_code_for_paper("2111.06389")
get_code_for_paper("2111.06389")
have("2111.06389")

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

curl https://syntology.ai/api/ran/2111.06389.json

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

[![Syntology run record](https://syntology.ai/api/ran/2111.06389.svg)](https://syntology.ai/paper/2111.06389)

Connect an agent — have() is free.