SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2310.01356 · 2023

Less is More: Toward Zero-Shot Local Scene Graph Generation via Foundation Models

arXiv · PDF · Open in the Atlas

Code that ran

6 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 6 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/2310.01356.json · badge: /api/ran/2310.01356.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
bowen-upenn/Multi-Agent-VQA community (classifier label)
classifier basis not recorded on the link
6 ran · 1 no recorded run
FunctionStatusWhere it lives
calculate_iou_batch Ran
ran on a synthesized input
bowen-upenn/Multi-Agent-VQA/utils.py
code served (permissive licence) · get_code("99d84bd80c9edece")
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, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create two sets of bounding boxes in format (x1, y1, x2, y2)
# a: N=2 bounding boxes
a = torch.tensor([
    [0.0, 0.0, 2.0, 2.0],
    [1.0, 1.0, 3.0, 3.0]
], dtype=torch.float32)

# b: M=3 bounding boxes
b = torch.tensor([
    [0.5, 0.5, 2.5, 2.5],
    [1.5, 1.5, 3.5, 3.5],
    [4.0, 4.0, 5.0, 5.0]
], dtype=torch.float32)

ARGS = {
    "a": a,
    "b": b
}
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, 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
extract_image_id Ran
ran on a synthesized input
bowen-upenn/Multi-Agent-VQA/utils_func/find_matched_rest_val.py
code served (permissive licence) · get_code("b9983bc3e2610d1b")
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 int, ndim 0, dtype family int, 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

ARGS = {
    'image_path': 'dataset_image_001.jpg'
}
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 int, ndim 0, dtype family int, 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
filter_boxes_pytorch Ran
ran on a synthesized input
bowen-upenn/Multi-Agent-VQA/utils.py
code served (permissive licence) · get_code("e686d5d1729b9df1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create two sets of bounding boxes
# Each bounding box is represented as [x1, y1, x2, y2]
# a: 2 bounding boxes
a = torch.tensor([
    [0.0, 0.0, 10.0, 10.0],
    [5.0, 5.0, 15.0, 15.0]
], dtype=torch.float32)

# b: 3 bounding boxes
b = torch.tensor([
    [1.0, 1.0, 9.0, 9.0],
    [20.0, 20.0, 30.0, 30.0],
    [6.0, 6.0, 14.0, 14.0]
], dtype=torch.float32)

iou_threshold = 0.5

ARGS = {
    'a': a,
    'b': b,
    'iou_threshold': iou_threshold
}
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
load_answer_list Ran
ran on a synthesized input
bowen-upenn/Multi-Agent-VQA/utils.py
code served (permissive licence) · get_code("5639f51ad9d50828")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import json
import tempfile
import os

# Create a temporary JSON file with sample answer data
temp_dir = tempfile.mkdtemp()
answer_file = os.path.join(temp_dir, "answers.json")

sample_answers = [
    "yes",
    "no",
    "red",
    "blue",
    "cat",
    "dog",
    "1",
    "2",
    "3",
    "person"
]

with open(answer_file, 'w') as f:
    json.dump(sample_answers, f)

ARGS = {
    "file_path": answer_file
}
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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
query_clip_count Ran
ran on a synthesized input
bowen-upenn/Multi-Agent-VQA/counting.py
code served (permissive licence) · get_code("412476f92c11ca4e")
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. 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 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
import numpy as np
from PIL import Image

torch.manual_seed(0)

# Create a minimal clip_count model that satisfies the function's usage
class MinimalClipCount(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(3, 4)
    
    def forward(self, image, prompts):
        # image shape: (batch, 3, 224, 224)
        # Should return a tuple where first element is the output
        batch_size = image.shape[0]
        # Return tensor of shape matching what draw_heat_map expects
        output = torch.randn(batch_size, 1, 384, 384)
        return (output,)

# Create a minimal image as numpy array (H, W, C)
image_array = np.random.randint(0, 256, (384, 384, 3), dtype=np.uint8)

# Create prompts list
prompts = ["a person", "multiple people"]

# Create the model
clip_count_model = MinimalClipCount()

# Device (CPU only)
device = torch.device('cpu')

ARGS = {
    'device': device,
    'image': image_array,
    'clip_count': clip_count_model,
    'prompts': prompts,
    'verbose': False,
    'save_heat_map': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, cv2, numpy, torch, torchvision — 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
query_sam Ran
ran on a synthesized input
bowen-upenn/Multi-Agent-VQA/detections.py
code served (permissive licence) · get_code("7eacd2e6da39e42f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from unittest.mock import Mock, MagicMock

torch.manual_seed(0)

# Create a mock SAM mask generator
sam_mask_generator = Mock()

# Create mock masks that would be returned by generate()
mock_masks = [
    {
        'segmentation': np.random.randint(0, 2, (100, 100), dtype=np.uint8),
        'area': 5000,
        'bbox': [10, 20, 50, 60],
        'predicted_iou': 0.95,
        'point_coords': np.array([[25.0, 40.0]]),
        'stability_score': 0.92,
        'crop_box': [0, 0, 100, 100]
    },
    {
        'segmentation': np.random.randint(0, 2, (100, 100), dtype=np.uint8),
        'area': 3000,
        'bbox': [30, 50, 40, 30],
        'predicted_iou': 0.88,
        'point_coords': np.array([[50.0, 65.0]]),
        'stability_score': 0.85,
        'crop_box': [0, 0, 100, 100]
    }
]

sam_mask_generator.generate = Mock(return_value=mock_masks)

# Create a mock image (RGB image as numpy array)
image = np.random.randint(0, 256, (100, 100, 3), dtype=np.uint8)

# Create a mock show_anns function to handle the visualization call
def show_anns(masks):
    pass

# Patch the show_anns in the module scope
import sys
current_module = sys.modules[__name__]
setattr(current_module, 'show_anns', show_anns)

# Define ARGS with all parameters
ARGS = {
    'device': 'cpu',
    'args': {},
    'sam_mask_generator': sam_mask_generator,
    'image': image
}
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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
draw_heat_map Not yet run
unverified (harvested, no recorded run)
bowen-upenn/Multi-Agent-VQA/counting.py
code served (permissive licence) · get_code("1a36aeec2b00943c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 7 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

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

Abstract

Humans inherently recognize objects via selective visual perception, transform specific regions from the visual field into structured symbolic knowledge, and reason their relationships among regions based on the allocation of limited attention resources in line with humans' goals. While it is intuitive for humans, contemporary perception systems falter in extracting structural information due to the intricate cognitive abilities and commonsense knowledge required. To fill this gap, we present a new task called Local Scene Graph Generation. Distinct from the conventional scene graph generation task, which encompasses generating all objects and relationships in an image, our proposed task aims to abstract pertinent structural information with partial objects and their relationships for boosting downstream tasks that demand advanced comprehension and reasoning capabilities. Correspondingly, we introduce zEro-shot Local scEne GrAph geNeraTion (ELEGANT), a framework harnessing foundation models renowned for their powerful perception and commonsense reasoning, where collaboration and information communication among foundation models yield superior outcomes and realize zero-shot local scene graph generation without requiring labeled supervision. Furthermore, we propose a novel open-ended evaluation metric, Entity-level CLIPScorE (ECLIPSE), surpassing previous closed-set evaluation metrics by transcending their limited label space, offering a broader assessment. Experiment results show that our approach markedly outperforms baselines in the open-ended evaluation setting, and it also achieves a significant performance boost of up to 24.58% over prior methods in the close-set setting, demonstrating the effectiveness and powerful reasoning ability of our proposed framework.

For agents

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

get_harvested_code_for_paper("2310.01356")
get_code_for_paper("2310.01356")
have("2310.01356")

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

curl https://syntology.ai/api/ran/2310.01356.json

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

[![Syntology run record](https://syntology.ai/api/ran/2310.01356.svg)](https://syntology.ai/paper/2310.01356)

Connect an agent — have() is free.