SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2207.05053 · 2022

Learning Continuous Grasping Function with a Dexterous Hand from Human Demonstrations

arXiv · PDF · Open in the Atlas

Code that ran

11 of 12 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; 2 honoured the contract we drafted, 0 violated it and 7 have no contract verdict; 2 failures of our instrument; 1 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/2207.05053.json · badge: /api/ran/2207.05053.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
jianglongye/cgf canonical (classifier label)
classifier basis not recorded on the link
11 ran · 1 no recorded run
FunctionStatusWhere it lives
axis_angle_to_matrix Ran
ran, honoured the contract we drafted
jianglongye/cgf/cgf/transformation.py
code served (permissive licence) · get_code("5d0ef4b4b1f7ecce")
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, 3, 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
import numpy as np

torch.manual_seed(0)

# Test case 1: torch.Tensor input with batch dimension
axis_angle_torch = torch.randn(2, 3)

# Test case 2: numpy array input with batch dimension
axis_angle_numpy = np.random.randn(2, 3)

# Test case 3: list input
axis_angle_list = [[0.1, 0.2, 0.3], [0.4, 0.5, 0.6]]

# Test case 4: single axis angle (3,) shape
axis_angle_single = torch.randn(3)

# Primary test case: torch tensor with batch
ARGS = {
    "axis_angle": torch.randn(2, 3)
}
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 Tensor, ndim 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 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
axis_angle_to_quaternion Ran
ran on a synthesized input
jianglongye/cgf/cgf/transformation.py
code served (permissive licence) · get_code("f6f74e1ef159b7ec")
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
import numpy as np

torch.manual_seed(0)

# Test with torch.Tensor
axis_angle_torch = torch.randn(2, 3)

ARGS = {
    "axis_angle": axis_angle_torch
}
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 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
batched_index_select Ran
ran; our contract draft was wrong, not the code
jianglongye/cgf/cgf/robotics.py
pointer only (licence: MIT); the stored copy get_code serves, from florasion/D-VQVAE, is pointer only (licence: NONE) · get_code("7d2905265c91d217")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from florasion/D-VQVAE, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

feature_transform_reguliarzer Ran
ran on a synthesized input
jianglongye/cgf/cgf/models.py
code served (permissive licence) · get_code("f3fc597a13ea9c07")
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. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 a transformation tensor
# The function expects: trans.size()[1] to get dimension d
# trans.is_cuda to check device
# torch.bmm(trans, trans.transpose(2, 1)) for batch matrix multiply
# So trans should be shape (batch_size, d, d) for a square transformation matrix

batch_size = 2
d = 4
trans = torch.randn(batch_size, d, d)

ARGS = {
    "trans": trans
}
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
load_joint_limits Ran
ran on a synthesized input
jianglongye/cgf/cgf/robotics.py
code served (permissive licence) · get_code("2d7565c102266a1f")
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 xml.etree.ElementTree as ET

# Create a temporary URDF file with valid joint limits
urdf_content = """<?xml version="1.0"?>
<robot name="test_robot">
  <joint name="joint_1" type="revolute">
    <limit lower="-1.57" upper="1.57" effort="10" velocity="1"/>
  </joint>
  <joint name="joint_2" type="prismatic">
    <limit lower="0.0" upper="0.5" effort="5" velocity="0.1"/>
  </joint>
  <joint name="joint_3" type="fixed">
    <limit lower="0.0" upper="0.0" effort="0" velocity="0"/>
  </joint>
  <joint name="joint_4" type="continuous">
    <limit lower="-3.14" upper="3.14" effort="20" velocity="2"/>
  </joint>
</robot>
"""

# Write URDF to a temporary file
with tempfile.NamedTemporaryFile(mode='w', suffix='.urdf', delete=False) as f:
    f.write(urdf_content)
    urdf_path = f.name

ARGS = {
    'urdf_path': urdf_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — 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
pc_normalize Ran
ran, honoured the contract we drafted
jianglongye/cgf/cgf/models.py
pointer only (licence: MIT); the stored copy get_code serves, from aboulch/ConvPoint, is pointer only (licence: NOASSERTION) · get_code("ec413739d406e611")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from aboulch/ConvPoint, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

rand_indices Ran
ran on a synthesized input
jianglongye/cgf/cgf/data.py
code served (permissive licence) · get_code("d951b12e257e2c90")
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 int; 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 numpy as np

# Set random seed for reproducibility
np.random.seed(0)

# Construct arguments for rand_indices function
# The function takes:
# - size: an integer representing the population size to sample from
# - sample_num: an integer representing how many samples to draw
# - is_sorted: a boolean indicating whether to sort the result

ARGS = {
    "size": 10,
    "sample_num": 5,
    "is_sorted": True
}
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 list, ndim 1, dtype family int — 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
rot_from_connected_link Ran
ran on a synthesized input
jianglongye/cgf/hand_teleop/utils/model_utils.py
code served (permissive licence) · get_code("16511581c7678dc5")
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, 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 position vectors for parent and child links
pos_parent = np.array([1.0, 2.0, 3.0])
pos_child = np.array([0.0, 1.0, 2.0])

# Create a parent rotation matrix (3x3)
parent_rotation = np.eye(3)

# Create an up axis vector
up_axis = np.array([0.0, 1.0, 0.0])

ARGS = {
    "pos_parent": pos_parent,
    "pos_child": pos_child,
    "parent_rotation": parent_rotation,
    "up_axis": up_axis
}
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, 3) — shape (3, 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
sample Ran
ran on a synthesized input
jianglongye/cgf/cgf/data.py
code served (permissive licence) · get_code("94d3f69b9739b003")
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 int, 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
import torch

torch.manual_seed(0)

# Create a small test array (numpy)
test_array_np = np.arange(20).reshape(4, 5)

# Create a small test tensor (torch)
test_tensor = torch.arange(20).reshape(4, 5)

ARGS = {
    "array": test_array_np,
    "sample_num": 3,
    "dim": 0,
    "is_sorted": True,
}
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 ndarray, ndim 2, dtype family int, 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
split_axis_angle Ran
ran on a synthesized input
jianglongye/cgf/cgf/transformation.py
code served (permissive licence) · get_code("7597de8adb8b3252")
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
import torch

torch.manual_seed(0)

# Test with numpy array input
axis_angle_np = np.array([[1.0, 0.0, 0.0, 0.5], [0.0, 1.0, 0.0, 0.3]])

# Test with torch tensor input
axis_angle_torch = torch.tensor([[1.0, 0.0, 0.0, 0.5], [0.0, 1.0, 0.0, 0.3]])

# Test with list input
axis_angle_list = [[1.0, 0.0, 0.0, 0.5], [0.0, 1.0, 0.0, 0.3]]

# Test with tuple input
axis_angle_tuple = ((1.0, 0.0, 0.0, 0.5), (0.0, 1.0, 0.0, 0.3))

ARGS = {
    "axis_angle": axis_angle_np
}
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 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
square_distance Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jianglongye/cgf/cgf/models.py
code served (permissive licence) · get_code("a3049715bcbcb43c")
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, 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 torch

torch.manual_seed(0)

# Create small tensors for src and dst
# src: [B, N, C] where B=2 (batch), N=3 (source points), C=4 (coordinates)
# dst: [B, M, C] where B=2 (batch), M=5 (target points), C=4 (coordinates)
src = torch.randn(2, 3, 4)
dst = torch.randn(2, 5, 4)

ARGS = {
    "src": src,
    "dst": dst,
}
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, 3, 5) — shape (2, 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
pt3d_knn Not yet run
unverified (harvested, no recorded run)
jianglongye/cgf/cgf/robotics.py
code served (permissive licence) · get_code("70486b666262f6d2")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 4 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 2 the same month, 4 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
square_distance 75 other papers
17 dated earlier, 3 the same month, 54 later, 1 same year, order not determinable · 9 only as an assembly brick
1706.02413 (2017-06)
PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
1811.07246 (2018-11)
PointConv: Deep Convolutional Networks on 3D Point Clouds
2003.00410 (2020-03)
PF-Net: Point Fractal Network for 3D Point Cloud Completion
and 72 more not named here
pc_normalize 36 other papers
14 dated earlier, 1 the same month, 21 later
1803.05827 (2018-03)
Local Spectral Graph Convolution for Point Set Feature Learning
1904.02375 (2019-04)
ConvPoint: Continuous Convolutions for Point Cloud Processing
1904.10014 (2019-04)
Linked Dynamic Graph CNN: Learning on Point Cloud via Linking Hierarchical Features
and 33 more not named here
axis_angle_to_matrix 3 other papers
1 dated earlier, 2 later · 1 only as an assembly brick
2203.04810 (2022-03)
Benchmarking Graphormer on Large-Scale Molecular Modeling Datasets
2210.01776 (2022-10)
DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking
2502.02513 (2025-02)
Diffusion Generative Modeling on Lie Group Representations
batched_index_select 1 other paper
1 later
2407.14062 (2024-07)
Decomposed Vector-Quantized Variational Autoencoder for Human Grasp 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

We propose to learn to generate grasping motion for manipulation with a dexterous hand using implicit functions. With continuous time inputs, the model can generate a continuous and smooth grasping plan. We name the proposed model Continuous Grasping Function (CGF). CGF is learned via generative modeling with a Conditional Variational Autoencoder using 3D human demonstrations. We will first convert the large-scale human-object interaction trajectories to robot demonstrations via motion retargeting, and then use these demonstrations to train CGF. During inference, we perform sampling with CGF to generate different grasping plans in the simulator and select the successful ones to transfer to the real robot. By training on diverse human data, our CGF allows generalization to manipulate multiple objects. Compared to previous planning algorithms, CGF is more efficient and achieves significant improvement on success rate when transferred to grasping with the real Allegro Hand. Our project page is available at https://jianglongye.com/cgf .

For agents

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

get_harvested_code_for_paper("2207.05053")
get_code_for_paper("2207.05053")
have("2207.05053")

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

curl https://syntology.ai/api/ran/2207.05053.json

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

[![Syntology run record](https://syntology.ai/api/ran/2207.05053.svg)](https://syntology.ai/paper/2207.05053)

Connect an agent — have() is free.