SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2503.00533 · ICLR · 2025

BodyGen: Advancing Towards Efficient Embodiment Co-Design

Zhe Li, Jianshu Li, Ruoyu Li, Junliang Xing, Haofei Lu, Zhe Wu, Yuanchun Shi, Mosat Mosat, Mosat

arXiv · PDF · Open in the Atlas

Code that ran

9 of 10 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 8 have no contract verdict; 0 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/2503.00533.json · badge: /api/ran/2503.00533.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
GenesisOrigin/BodyGen community (classifier label)
classifier basis not recorded on the link
9 ran · 1 no recorded run
FunctionStatusWhere it lives
array_to_str Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/utils/tools.py
code served (permissive licence) · get_code("49fb2c1c4bf2c4d7")
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 torch
import numpy as np

torch.manual_seed(0)

# Analyzing the function:
# - arr: needs to support len() and be iterable with values that can be formatted
# - format_str: a format string that will be repeated len(arr) times
# - sep_str: a string used as separator (default ' ')

# Create a simple list of numbers that can be formatted
arr = [1, 2, 3, 4]

# Format string that expects a single numeric argument
format_str = "{:.2f}"

# Separator string
sep_str = " | "

ARGS = {
    'arr': arr,
    'format_str': format_str,
    'sep_str': sep_str
}
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 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
create_logger Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/utils/logger.py
code served (permissive licence) · get_code("e80e9b137849e33a")
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 Logger. 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 logging
import os
import tempfile

# Create a temporary directory for the log file
temp_dir = tempfile.mkdtemp()
log_filename = os.path.join(temp_dir, "test_logger.log")

ARGS = {
    "filename": log_filename,
    "file_handle": True,
}
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 Logger — a Logger output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Logger 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_eta_str Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/utils/tools.py
code served (permissive licence) · get_code("de610578837875a2")
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 datetime

def get_eta_str(cur_iter, total_iter, time_per_iter):
    eta = time_per_iter * (total_iter - cur_iter - 1)
    return str(datetime.timedelta(seconds=round(eta)))

ARGS = {
    "cur_iter": 42,
    "total_iter": 100,
    "time_per_iter": 1.5
}
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 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
get_single_body_qposaddr Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/utils/mujoco.py
code served (permissive licence) · get_code("8dca878ea5e97bf4")
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, 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

class MockMujocoModel:
    def __init__(self):
        # body_names: list of body names
        self.body_names = ['world', 'torso', 'left_arm', 'right_arm']
        
        # body_jntadr: index into joint arrays for each body
        # body 0 (world): joint address 0
        # body 1 (torso): joint address 1
        # body 2 (left_arm): joint address 3
        # body 3 (right_arm): joint address 5
        self.body_jntadr = np.array([0, 1, 3, 5], dtype=np.int32)
        
        # body_jntnum: number of joints for each body
        # world: 0 joints
        # torso: 2 joints
        # left_arm: 2 joints
        # right_arm: 1 joint
        self.body_jntnum = np.array([0, 2, 2, 1], dtype=np.int32)
        
        # jnt_qposadr: qpos address for each joint
        # joint 0: qpos addr 0
        # joint 1: qpos addr 0
        # joint 2: qpos addr 3
        # joint 3: qpos addr 6
        # joint 4: qpos addr 7
        # joint 5: qpos addr 8
        self.jnt_qposadr = np.array([0, 0, 3, 6, 7, 8], dtype=np.int32)
        
        # nq: total number of qpos
        self.nq = 9

model = MockMujocoModel()
body = 'left_arm'

ARGS = {
    'model': model,
    'body': body,
}
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, ndim 1, dtype family int — a tuple 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
normal_entropy Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/utils/math.py
code served (permissive licence) · get_code("8eee684b0bb59dbf")
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, 1) 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 math

torch.manual_seed(0)

# The function expects a tensor `std` with shape (batch_size, dim)
# It calls .pow(2) on it, torch.log on the result, and .sum(1, keepdim=True)
# So we need a 2D tensor where dim 0 is batch and dim 1 is the feature dimension

std = torch.randn(2, 4).abs() + 0.1  # batch=2, features=4, ensure positive values

ARGS = {
    "std": std
}
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, 1) — shape (2, 1) 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
normal_log_density Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/utils/math.py
code served (permissive licence) · get_code("b9112c793778d273")
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, 1) 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 math

torch.manual_seed(0)

# Create small tensors for testing
# x: input tensor, shape (batch_size, dim)
x = torch.randn(2, 4)

# mean: mean of the normal distribution, shape (batch_size, dim)
mean = torch.randn(2, 4)

# log_std: log of standard deviation, shape (batch_size, dim)
log_std = torch.randn(2, 4)

# std: standard deviation, shape (batch_size, dim)
std = torch.exp(log_std)

ARGS = {
    "x": x,
    "mean": mean,
    "log_std": log_std,
    "std": std
}
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, 1) — shape (2, 1) 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
normalize_range Ran
ran, honoured the contract we drafted
GenesisOrigin/BodyGen/khrylib/robot/xml_robot.py
code served (permissive licence) · get_code("66a0415bf6172ec1")
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 numpy as np

torch.manual_seed(0)

# Analyze the function:
# normalize_range(value, lb, ub)
# - value: numeric value to normalize
# - lb: lower bound
# - ub: upper bound
# The function performs linear normalization to [-1, 1] range

# Create test values - using scalars and tensors for variety
value = torch.tensor(5.0)
lb = torch.tensor(0.0)
ub = torch.tensor(10.0)

ARGS = {
    "value": value,
    "lb": lb,
    "ub": ub
}
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 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
parse_fromto Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/robot/xml_robot.py
code served (permissive licence) · get_code("9ce3c0fc66cce890")
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, ndim 2, 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 numpy as np

def parse_fromto(string):
    fromto = np.fromstring(string, sep=' ')
    return (fromto[:3], fromto[3:])

# The function takes a single parameter 'string' which should be a space-separated
# string of numbers that can be parsed by np.fromstring
# It splits the resulting array into two parts: first 3 elements and rest

ARGS = {
    'string': '1.0 2.0 3.0 4.0 5.0 6.0'
}
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, ndim 2, dtype family float — a tuple 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
parse_vec Ran
ran on a synthesized input
GenesisOrigin/BodyGen/khrylib/robot/xml_robot.py
code served (permissive licence) · get_code("96ecda6cfeb2c3b1")
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 1, dtype family float, shape (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

def parse_vec(string):
    return np.fromstring(string, sep=' ')

# Construct a valid string argument with space-separated numbers
string_arg = "1.5 2.3 3.7 4.1 5.9"

ARGS = {
    "string": string_arg
}
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 1, dtype family float, shape (5,) — shape (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_img Not yet run
unverified (harvested, no recorded run)
GenesisOrigin/BodyGen/khrylib/utils/tools.py
code served (permissive licence) · get_code("5726721449c916d8")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 7 are also present in other papers whose code we hold (7 with at least one paper dated earlier, 0 the same month, 0 later) and 3 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
create_logger 4 other papers
4 dated earlier
2108.08422 (2021-08)
Generating Smooth Pose Sequences for Diverse Human Motion Prediction
2110.03659 (2021-10)
Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design
aaai_25754 (2023 (year only))
Human Joint Kinematics Diffusion-Refinement for Stochastic Motion Prediction
and 1 more not named here
normal_entropy 2 other papers
2 dated earlier
2110.03659 (2021-10)
Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design
2306.00036 (2023-06)
Symmetry-Aware Robot Design with Structured Subgroups
normal_log_density 2 other papers
2 dated earlier
2110.03659 (2021-10)
Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design
2306.00036 (2023-06)
Symmetry-Aware Robot Design with Structured Subgroups
normalize_range 2 other papers
2 dated earlier
2110.03659 (2021-10)
Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design
2306.00036 (2023-06)
Symmetry-Aware Robot Design with Structured Subgroups
parse_fromto 2 other papers
2 dated earlier
2110.03659 (2021-10)
Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design
2306.00036 (2023-06)
Symmetry-Aware Robot Design with Structured Subgroups
parse_vec 2 other papers
2 dated earlier
2110.03659 (2021-10)
Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design
2306.00036 (2023-06)
Symmetry-Aware Robot Design with Structured Subgroups
get_single_body_qposaddr 1 other paper
1 dated earlier
2110.03659 (2021-10)
Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design

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

Embodiment co-design aims to optimize a robot's morphology and control policy simultaneously. While prior work has demonstrated its potential for generating environment-adaptive robots, this field still faces persistent challenges in optimization efficiency due to the (i) combinatorial nature of morphological search spaces and (ii) intricate dependencies between morphology and control. We prove that the ineffective morphology representation and unbalanced reward signals between the design and control stages are key obstacles to efficiency. To advance towards efficient embodiment co-design, we propose BodyGen, which utilizes (1) topology-aware self-attention for both design and control, enabling efficient morphology representation with lightweight model sizes; (2) a temporal credit assignment mechanism that ensures balanced reward signals for optimization. With our findings, Body achieves an average 60.03% performance improvement against state-of-the-art baselines. We provide codes and more results on the website: https://genesisorigin.github.io.

For agents

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

get_harvested_code_for_paper("2503.00533")
get_code_for_paper("2503.00533")
have("2503.00533")

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

curl https://syntology.ai/api/ran/2503.00533.json

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

[![Syntology run record](https://syntology.ai/api/ran/2503.00533.svg)](https://syntology.ai/paper/2503.00533)

Connect an agent — have() is free.