SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2312.01697 · 2023

Hulk: A Universal Knowledge Translator for Human-Centric Tasks

arXiv · PDF · Open in the Atlas

Code that ran

13 of 24 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 1 has no contract verdict; 11 failures of our instrument; 11 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/2312.01697.json · badge: /api/ran/2312.01697.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
opengvlab/hulk canonical (classifier label)
classifier basis: census
12 ran · 7 no recorded run
opengvlab/humanbench canonical (classifier label)
classifier basis: census
1 ran · 4 no recorded run
FunctionStatusWhere it lives
accuracy Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
opengvlab/hulk/core/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("78d1f9f766be74b7")
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 luomingshuang/M3GPT, 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.

accuracy_multi Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("8a63c5da44344ba6")
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 luomingshuang/M3GPT, 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.

all_gather Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/comm_.py
pointer only (licence: MIT); the stored copy get_code serves, from vis-nlp/ChartQA, is pointer only (licence: GPL-3.0) · get_code("41c46b0d5713eb14")
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 vis-nlp/ChartQA, and GPL-3.0, as GitHub detects it on that repository, is not in the permissive set Syntology serves inline. 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.

allreduce Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
opengvlab/hulk/core/distributed_utils.py
code served (permissive licence) · get_code("ae5f2eed69658ac0")
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 torch.distributed as dist
from unittest.mock import MagicMock, patch

torch.manual_seed(0)

# Mock dist.all_reduce to avoid requiring actual distributed setup
original_all_reduce = dist.all_reduce

def mock_all_reduce(x, group=None):
    """Mock implementation that just returns the tensor unchanged"""
    return x

# Patch dist.all_reduce
dist.all_reduce = mock_all_reduce

# Create a small tensor for the function to process
x = torch.randn(2, 4)

# group_idx can be None or 0 (which gets converted to None in the function)
group_idx = 0

ARGS = {
    "x": x,
    "group_idx": group_idx,
}
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
broadcast_object Ran
ran on a synthesized input
opengvlab/hulk/core/comm_.py
code served (permissive licence) · get_code("c632684a9726a77a")
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 torch.distributed as dist
from unittest.mock import MagicMock, patch

torch.manual_seed(0)

# Mock the distributed module since we're in a single-process environment
# and can't actually initialize a process group
with patch('torch.distributed.broadcast_object_list') as mock_broadcast:
    # Set up the mock to simulate broadcast behavior
    def broadcast_side_effect(data_list, src=0, group=None):
        # In a real scenario, this would broadcast the first element to all ranks
        # For testing, we just keep the list as-is
        pass
    
    mock_broadcast.side_effect = broadcast_side_effect

# Test data - various picklable objects
test_string = "hello world"
test_dict = {"key": "value", "number": 42}
test_list = [1, 2, 3, 4, 5]
test_tensor = torch.randn(2, 4)

# Create a mock process group (not actually used in the function body,
# but accepted as a parameter)
mock_group = MagicMock()

ARGS = {
    "data": test_string,
    "src": 0,
    "group": None,
}
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 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
dist_init Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
opengvlab/humanbench/PATH/core/distributed_utils.py
code served (permissive licence) · get_code("5d01dded5497144b")
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. 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 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 os
import torch
import torch.distributed as dist

# Set up minimal SLURM environment variables
os.environ['SLURM_PROCID'] = '0'
os.environ['SLURM_NTASKS'] = '1'
os.environ['SLURM_NODELIST'] = 'localhost'

# Since this function calls dist.init_process_group which requires actual distributed setup,
# we'll mock it to avoid needing a real distributed environment
original_init = dist.init_process_group

def mock_init_process_group(*args, **kwargs):
    """Mock the distributed initialization to avoid needing real distributed setup"""
    pass

dist.init_process_group = mock_init_process_group

# Also mock subprocess.getoutput to return a valid hostname
import subprocess
original_getoutput = subprocess.getoutput

def mock_getoutput(cmd):
    """Mock scontrol command to return localhost"""
    if 'scontrol' in cmd:
        return 'localhost'
    return original_getoutput(cmd)

subprocess.getoutput = mock_getoutput

ARGS = {
    'method': 'slurm',
    'port': '5671'
}
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, 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
flat Ran
ran, honoured the contract we drafted
opengvlab/hulk/core/config.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("4570d6d70bccdfb6")
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 luomingshuang/M3GPT, 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.

gather Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/comm_.py
pointer only (licence: MIT); the stored copy get_code serves, from vis-nlp/ChartQA, is pointer only (licence: GPL-3.0) · get_code("00a9f7918ac39329")
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 vis-nlp/ChartQA, and GPL-3.0, as GitHub detects it on that repository, is not in the permissive set Syntology serves inline. 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.

init Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/fp16/amp.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("eb8a2600afc1de95")
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 luomingshuang/M3GPT, 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.

param_group_fc Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/make_param_group.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("70165e0988ce9571")
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 luomingshuang/M3GPT, 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.

param_group_multitask Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/make_param_group.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("e4a7663d686ce91a")
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 luomingshuang/M3GPT, 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.

param_group_no_wd Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/make_param_group.py
code served (permissive licence) · get_code("e4f5737bd720668d")
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
import torch.nn as nn
from torch.nn import BatchNorm2d, BatchNorm1d, SyncBatchNorm

torch.manual_seed(0)

# Create a small model with Conv2d, Linear, and BatchNorm layers
# to test the param_group_no_wd function
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 4, kernel_size=3, bias=True)
        self.bn1 = nn.BatchNorm2d(4)
        self.linear1 = nn.Linear(4, 2, bias=True)
        self.bn2 = nn.BatchNorm1d(2)
        self.linear2 = nn.Linear(2, 1, bias=False)
    
    def forward(self, x):
        return x

model = SimpleModel()

ARGS = {
    'model': model
}
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
retry_if_cuda_oom Ran
ran; our contract draft was wrong, not the code
opengvlab/hulk/core/memory.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("ff1253be17efe1ae")
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 luomingshuang/M3GPT, 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.

allreduce_async Not yet run
unverified (harvested, no recorded run)
opengvlab/hulk/core/distributed_utils.py
code served (permissive licence) · get_code("02640c5911ca6b56")
How it ran

no invocation recorded.

broadcast Not yet run
unverified (harvested, no recorded run)
opengvlab/hulk/core/distributed_utils.py
code served (permissive licence) · get_code("305673e9acaebcc6")
How it ran

no invocation recorded.

float_function Not yet run
unverified (harvested, no recorded run)
opengvlab/hulk/core/fp16/amp.py
code served (permissive licence) · get_code("f27c7ab614e20039")
How it ran

no invocation recorded.

half_function Not yet run
unverified (harvested, no recorded run)
opengvlab/hulk/core/fp16/amp.py
code served (permissive licence) · get_code("6b4eee4b4bb1f238")
How it ran

no invocation recorded.

load_last_iter Not yet run
unverified (harvested, no recorded run)
opengvlab/hulk/core/utils.py
code served (permissive licence) · get_code("7c26a4ba6aca54f0")
How it ran

no invocation recorded.

param_group_no_wd Not yet run
unverified (harvested, no recorded run)
opengvlab/humanbench/PATH/core/make_param_group.py
code served (permissive licence) · get_code("b14254d606cb6719")
How it ran

no invocation recorded.

simple_group_split Not yet run
unverified (harvested, no recorded run)
opengvlab/humanbench/PATH/core/distributed_utils.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("b1471eccec180d87")
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 luomingshuang/M3GPT, 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. No invocation recorded.

specific_group_split Not yet run
unverified (harvested, no recorded run)
opengvlab/humanbench/PATH/core/distributed_utils.py
pointer only (licence: MIT); the stored copy get_code serves, from luomingshuang/M3GPT, is pointer only (licence: NONE) · get_code("4e36f197c125d4d3")
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 luomingshuang/M3GPT, 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. No invocation recorded.

specific_group_split Not yet run
unverified (harvested, no recorded run)
opengvlab/hulk/core/config.py
code served (permissive licence) · get_code("831beea1b4db57e3")
How it ran

no invocation recorded.

specific_group_split Not yet run
unverified (harvested, no recorded run)
opengvlab/humanbench/PATH/core/config.py
code served (permissive licence) · get_code("62ad1cdb3bc504fe")
How it ran

no invocation recorded.

specific_group_split_modality_groups Not yet run
unverified (harvested, no recorded run)
opengvlab/hulk/core/config.py
code served (permissive licence) · get_code("6f5d8cf806cbe77c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 24 samples lifted from this paper's linked repositories, 24 carry a function identity and 0 do not. Of the 24 distinct functions with an identity, 14 are also present in other papers whose code we hold (14 with at least one paper dated earlier, 0 the same month, 13 later) and 10 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
all_gather 16 other papers
9 dated earlier, 4 later, 3 same year, order not determinable
2108.02833 (2021-08)
Elaborative Rehearsal for Zero-shot Action Recognition
136730727 (2022 (year only))
Self-Supervised Sparse Representation for Video Anomaly Detection
2023.findings-emnlp.644 (2023 (year only))
VIP5: Towards Multimodal Foundation Models for Recommendation
and 13 more not named here
gather 14 other papers
7 dated earlier, 4 later, 3 same year, order not determinable
2108.02833 (2021-08)
Elaborative Rehearsal for Zero-shot Action Recognition
136730727 (2022 (year only))
Self-Supervised Sparse Representation for Video Anomaly Detection
2023.findings-emnlp.644 (2023 (year only))
VIP5: Towards Multimodal Foundation Models for Recommendation
and 11 more not named here
accuracy 4 other papers
2 dated earlier, 2 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2307.09742 (2023-07)
Improved Distribution Matching for Dataset Condensation
2402.10635 (2024-02)
ContiFormer: Continuous-Time Transformer for Irregular Time Series Modeling
and 1 more not named here
float_function 3 other papers
2 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2310.18986 (2023-10)
Controllable Group Choreography using Contrastive Diffusion
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
half_function 3 other papers
2 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2310.18986 (2023-10)
Controllable Group Choreography using Contrastive Diffusion
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
accuracy_multi 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
flat 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
init 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
param_group_fc 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
param_group_multitask 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
retry_if_cuda_oom 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
simple_group_split 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
specific_group_split 2 other papers
1 dated earlier, 1 later
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions
2405.16273 (2024-05)
M 3 GPT: An Advanced Multimodal, Multitask Framework for Motion Comprehension and Generation
dist_init 1 other paper
1 dated earlier
2303.02936 (2023-03)
UniHCP: A Unified Model for Human-Centric Perceptions

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

Human-centric perception tasks, e.g., pedestrian detection, skeleton-based action recognition, and pose estimation, have wide industrial applications, such as metaverse and sports analysis. There is a recent surge to develop human-centric foundation models that can benefit a broad range of human-centric perception tasks. While many human-centric foundation models have achieved success, they did not explore 3D and vision-language tasks for human-centric and required task-specific finetuning. These limitations restrict their application to more downstream tasks and situations. To tackle these problems, we present Hulk, the first multimodal human-centric generalist model, capable of addressing 2D vision, 3D vision, skeleton-based, and vision-language tasks without task-specific finetuning. The key to achieving this is condensing various task-specific heads into two general heads, one for discrete representations, \emph{e.g.,} languages, and the other for continuous representations, \emph{e.g.,} location coordinates. The outputs of two heads can be further stacked into four distinct input and output modalities. This uniform representation enables Hulk to treat diverse human-centric tasks as modality translation, integrating knowledge across a wide range of tasks. Comprehensive evaluations of Hulk on 12 benchmarks covering 8 human-centric tasks demonstrate the superiority of our proposed method, achieving state-of-the-art performance in 11 benchmarks. The code will be available on https://github.com/OpenGVLab/Hulk.

For agents

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

get_harvested_code_for_paper("2312.01697")
get_code_for_paper("2312.01697")
have("2312.01697")

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

curl https://syntology.ai/api/ran/2312.01697.json

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

[![Syntology run record](https://syntology.ai/api/ran/2312.01697.svg)](https://syntology.ai/paper/2312.01697)

Connect an agent — have() is free.