SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2502.02673 · 2025

MedRAX: Medical Reasoning Agent for Chest X-ray

arXiv · PDF · Open in the Atlas

Code that ran

12 of 13 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, 1 violated it and 3 have no contract verdict; 6 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/2502.02673.json · badge: /api/ran/2502.02673.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
bowang-lab/medrax canonical (classifier label)
classifier basis not recorded on the link
12 ran · 1 no recorded run
FunctionStatusWhere it lives
compare_messages_gen Ran
ran; our contract draft was wrong, not the code
bowang-lab/medrax/medrax/llava/eval/eval_multimodal_chat_gpt_score.py
code served (permissive licence) · get_code("fcb74c8a86642376")
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

# Define all parameters for compare_messages_gen
ARGS = {
    'fig_label': 'Figure 1',
    'fig_caption': 'A cat sitting on a mat in a sunny room',
    'fig_context': 'Indoor scene with natural lighting',
    'question': 'What is the cat doing in this image?',
    'ans1': 'The cat is sitting on a mat and appears to be relaxing in the sunlight coming through the window.',
    'ans2': 'The cat is positioned on a mat in an indoor environment with natural light from a window.'
}
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
conv_to_str Ran
ran; our contract draft was wrong, not the code
bowang-lab/medrax/medrax/llava/eval/eval_multimodal_chat_gpt_score.py
code served (permissive licence) · get_code("c7eacf8297119313")
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

torch.manual_seed(0)

ARGS = {
    'fig_label': 'Figure 1',
    'fig_caption': 'A cat sitting on a windowsill looking outside at birds',
    'fig_context': 'Indoor scene with natural lighting from the window',
    'question': 'What is the cat looking at and why might it be interested?',
    'ans1': 'The cat is looking at the birds outside the window. Cats are naturally attracted to birds due to their hunting instincts, and the movement and sounds of birds capture their attention.',
    'ans2': 'The cat appears to be gazing out at the outdoor environment. Birds visible through the window attract feline interest because cats have predatory instincts that make moving objects particularly engaging.'
}
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
expand2square Ran
ran; our contract draft was wrong, not the code
bowang-lab/medrax/medrax/llava/mm_utils.py
code served (permissive licence) · get_code("908505ed68ff4871")
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 Image. 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.

from PIL import Image
import random

# Set seed for reproducibility
random.seed(0)

# Create a small test image (rectangular, not square)
# Using RGB mode with dimensions 100x60 (wider than tall)
test_image = Image.new('RGB', (100, 60), color='white')

# Define background color for padding
background_color = (128, 128, 128)  # Gray color

ARGS = {
    'pil_img': test_image,
    'background_color': background_color
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int 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_chunk Ran
ran; our contract draft was wrong, not the code
bowang-lab/medrax/medrax/llava/eval/model_vqa.py
code served (permissive licence) · get_code("42a46570620cd9fa")
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) 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 math

def split_list(lst, n):
    """Split a list into n (roughly) equal-sized chunks"""
    chunk_size = math.ceil(len(lst) / n)
    return [lst[i:i + chunk_size] for i in range(0, len(lst), chunk_size)]

def get_chunk(lst, n, k):
    chunks = split_list(lst, n)
    return chunks[k]

# Setup code
lst = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
n = 3
k = 1

ARGS = {
    "lst": lst,
    "n": n,
    "k": k
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 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
load_image_from_base64 Ran
ran, honoured the contract we drafted
bowang-lab/medrax/medrax/llava/mm_utils.py
code served (permissive licence) · get_code("c3ee9d07c900dd55")
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 PngImageFile. 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 base64
from PIL import Image
from io import BytesIO
import numpy as np

# Create a small valid image in memory
img_array = np.uint8(np.random.rand(10, 10, 3) * 255)
img = Image.fromarray(img_array, mode='RGB')

# Encode to base64
buffer = BytesIO()
img.save(buffer, format='PNG')
buffer.seek(0)
image_base64 = base64.b64encode(buffer.read()).decode('utf-8')

ARGS = {
    'image': image_base64
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type PngImageFile — a PngImageFile output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PngImageFile output the recorded ndim 3 and dtype family int 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
load_prompts_from_file Ran
ran on a synthesized input
bowang-lab/medrax/medrax/utils/utils.py
code served (permissive licence) · get_code("ac5d4e8611111ae8")
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 os
import tempfile

# Create a temporary file with prompts in the expected format
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write('[prompt1]\n')
temp_file.write('This is the first prompt content.\n')
temp_file.write('It can span multiple lines.\n')
temp_file.write('[prompt2]\n')
temp_file.write('This is the second prompt.\n')
temp_file.write('[prompt3]\n')
temp_file.write('Another prompt here.\n')
temp_file.close()

ARGS = {
    'file_path': temp_file.name
}
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
load_system_prompt Ran
ran on a synthesized input
bowang-lab/medrax/medrax/utils/utils.py
code served (permissive licence) · get_code("a76ec818e2e53ab4")
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 os
import json
import tempfile

# Create temporary files with the required structure
temp_dir = tempfile.mkdtemp()

# Create system prompts file
system_prompts_file = os.path.join(temp_dir, "system_prompts.txt")
with open(system_prompts_file, 'w') as f:
    f.write("[GENERAL_ASSISTANT]\n")
    f.write("You are a helpful assistant.\n")
    f.write("You can use various tools to help the user.\n")
    f.write("[EXPERT_MODE]\n")
    f.write("You are an expert assistant with advanced capabilities.\n")

# Create tools.json file
tools_json_path = os.path.join(temp_dir, "tools.json")
tools_data = {
    "calculator": {
        "description": "A tool for mathematical calculations",
        "prompt": "Use this tool to perform arithmetic operations",
        "input_type": "string",
        "return_type": "float"
    },
    "search": {
        "description": "A tool for searching information",
        "prompt": "Use this tool to search for information online",
        "input_type": "string",
        "return_type": "string"
    },
    "weather": {
        "description": "A tool for getting weather information",
        "prompt": "Use this tool to get current weather",
        "input_type": "string",
        "return_type": "dict"
    }
}
with open(tools_json_path, 'w') as f:
    json.dump(tools_data, f)

# Define arguments for load_system_prompt
ARGS = {
    "system_prompts_file": system_prompts_file,
    "system_prompt_type": "GENERAL_ASSISTANT",
    "tools": ["calculator", "search"],
    "tools_json_path": tools_json_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 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
load_tool_prompts Ran
ran on a synthesized input
bowang-lab/medrax/medrax/utils/utils.py
code served (permissive licence) · get_code("c097d8eeb37f0b88")
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 os
import json
import tempfile

# Create a temporary directory and tools.json file for testing
temp_dir = tempfile.mkdtemp()
tools_json_path = os.path.join(temp_dir, 'tools.json')

# Create sample tools data
tools_data = {
    "calculator": {
        "description": "A tool for performing mathematical calculations",
        "prompt": "Use this tool to calculate mathematical expressions",
        "input_type": "string",
        "return_type": "float"
    },
    "search": {
        "description": "A tool for searching information",
        "prompt": "Use this tool to search for information on the web",
        "input_type": "string",
        "return_type": "list"
    },
    "weather": {
        "description": "A tool for checking weather",
        "prompt": "Use this tool to get weather information",
        "input_type": "location",
        "return_type": "dict"
    }
}

# Write the tools data to the JSON file
with open(tools_json_path, 'w') as f:
    json.dump(tools_data, f)

# Define the arguments for the function
ARGS = {
    'tools': ['calculator', 'search'],
    'tools_json_path': tools_json_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 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
process_images Ran
ran; our contract draft was wrong, not the code
bowang-lab/medrax/medrax/llava/mm_utils.py
code served (permissive licence) · get_code("f0f8c9c5a77c9b5b")
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 4, dtype family float, shape (2, 3, 4, 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
from PIL import Image
import random

torch.manual_seed(0)
random.seed(0)

# Create a stand-in image processor with the required attributes and methods
class StandInImageProcessor:
    def __init__(self):
        self.image_mean = [0.5, 0.5, 0.5]  # for RGB images
    
    def preprocess(self, image, return_tensors=None):
        # Convert PIL image to tensor of shape (3, H, W)
        img_array = torch.randn(3, 4, 4)  # tiny tensor (3 channels, 4x4 spatial)
        if return_tensors == 'pt':
            return {'pixel_values': img_array.unsqueeze(0)}  # Add batch dimension
        return img_array

# Create a stand-in model config with the required attributes
class StandInModelConfig:
    def __init__(self):
        self.image_aspect_ratio = 'pad'

# Create sample PIL images (RGB mode, different sizes)
image1 = Image.new('RGB', (5, 3), color=(100, 150, 200))
image2 = Image.new('RGB', (5, 3), color=(50, 75, 100))
images = [image1, image2]

# Create instances
image_processor = StandInImageProcessor()
model_cfg = StandInModelConfig()

ARGS = {
    'images': images,
    'image_processor': image_processor,
    'model_cfg': model_cfg
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float, shape (2, 3, 4, 4) — shape (2, 3, 4, 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
split_list Ran
ran; our contract draft was wrong, not the code
bowang-lab/medrax/medrax/llava/eval/model_vqa.py
code served (permissive licence) · get_code("076c252c52cbb161")
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) and returned list. 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 math

def split_list(lst, n):
    """Split a list into n (roughly) equal-sized chunks"""
    chunk_size = math.ceil(len(lst) / n)
    return [lst[i:i + chunk_size] for i in range(0, len(lst), chunk_size)]

# Setup code
lst = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
n = 3

ARGS = {
    "lst": lst,
    "n": n
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 — a list 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
sum_list_list Ran
ran, honoured the contract we drafted
bowang-lab/medrax/medrax/llava/eval/eval_multimodal_chat_gpt_score.py
code served (permissive licence) · get_code("90719ea4ae0ed586")
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
import numpy as np

torch.manual_seed(0)

# The function sum_list_list takes x which should be a list of lists
# It iterates through inner lists and sums all individual items
# The items appear to be numeric values that support the sum() operation

# Create a list of lists with numeric values
x = [
    [1, 2, 3, 4],
    [5, 6],
    [7, 8, 9]
]

ARGS = {
    "x": x
}
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 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
violates_moderation Ran
ran, violated the contract we drafted
bowang-lab/medrax/medrax/llava/utils.py
code served (permissive licence) · get_code("d4ba997957e75202")
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 bool, 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 os

# Set a dummy API key to avoid KeyError when accessing os.environ['OPENAI_API_KEY']
os.environ['OPENAI_API_KEY'] = 'sk-test-dummy-key-for-testing'

# The function takes a single parameter: text
# It expects a string that will be sent to the OpenAI moderation API
text_input = "This is a sample text to check for moderation violations."

ARGS = {
    'text': text_input
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
requests — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, 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
pretty_print_semaphore Not yet run
unverified (harvested, no recorded run)
bowang-lab/medrax/medrax/llava/utils.py
code served (permissive licence) · get_code("37899f22fb191b37")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 distinct functions with an identity, 10 are also present in other papers whose code we hold (9 with at least one paper dated earlier, 3 the same month, 10 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
split_list 197 other papers
153 dated earlier, 5 the same month, 31 later, 7 same year, order not determinable, 1 undated
2112.05682 (2021-12)
Self-attention Does Not Need $O(n^2)$ Memory
2306.17107 (2023-06)
LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding
2307.10490 (2023-07)
Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs
and 194 more not named here
get_chunk 189 other papers
148 dated earlier, 4 the same month, 30 later, 7 same year, order not determinable
2112.05682 (2021-12)
Self-attention Does Not Need $O(n^2)$ Memory
2306.17107 (2023-06)
LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image Understanding
2307.10490 (2023-07)
Abusing Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs
and 186 more not named here
load_image_from_base64 85 other papers
72 dated earlier, 1 the same month, 11 later, 1 same year, order not determinable
2112.05682 (2021-12)
Self-attention Does Not Need $O(n^2)$ Memory
2308.10253 (2023-08)
StableLLaVA: Enhanced Visual Instruction Tuning with Synthesized Image-Dialogue Data
2309.07120 (2023-09)
Sight Beyond Text: Multi-Modal Training Enhances LLMs in Truthfulness and Ethics
and 82 more not named here
pretty_print_semaphore 54 other papers
45 dated earlier, 7 later, 2 same year, order not determinable
2106.06909 (2021-06)
GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio
2308.13437 (2023-08)
Position-Enhanced Visual Instruction Tuning for Multimodal Large Language Models
2309.07120 (2023-09)
Sight Beyond Text: Multi-Modal Training Enhances LLMs in Truthfulness and Ethics
and 51 more not named here
compare_messages_gen 3 other papers
1 dated earlier, 2 later
2406.13173 (2024-06)
Biomedical Visual Instruction Tuning with Clinician Preference Alignment
2601.17918 (2026-01)
Benchmarking Direct Preference Optimization for Medical Large Vision-Language Models
2605.27378 (2026-05)
OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis
conv_to_str 3 other papers
1 dated earlier, 2 later
2406.13173 (2024-06)
Biomedical Visual Instruction Tuning with Clinician Preference Alignment
2601.17918 (2026-01)
Benchmarking Direct Preference Optimization for Medical Large Vision-Language Models
2605.27378 (2026-05)
OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis
expand2square 3 other papers
1 dated earlier, 2 later
2406.13173 (2024-06)
Biomedical Visual Instruction Tuning with Clinician Preference Alignment
2601.17918 (2026-01)
Benchmarking Direct Preference Optimization for Medical Large Vision-Language Models
2605.27378 (2026-05)
OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis
process_images 3 other papers
1 dated earlier, 2 later
2406.13173 (2024-06)
Biomedical Visual Instruction Tuning with Clinician Preference Alignment
2601.17918 (2026-01)
Benchmarking Direct Preference Optimization for Medical Large Vision-Language Models
2605.27378 (2026-05)
OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis
sum_list_list 3 other papers
1 dated earlier, 2 later
2406.13173 (2024-06)
Biomedical Visual Instruction Tuning with Clinician Preference Alignment
2601.17918 (2026-01)
Benchmarking Direct Preference Optimization for Medical Large Vision-Language Models
2605.27378 (2026-05)
OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis
violates_moderation 1 other paper
1 later
2605.27378 (2026-05)
OralAgent: Integrating Reasoning, Tools, and Knowledge for Interactive Dental Image Analysis

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

Chest X-rays (CXRs) play an integral role in driving critical decisions in disease management and patient care. While recent innovations have led to specialized models for various CXR interpretation tasks, these solutions often operate in isolation, limiting their practical utility in clinical practice. We present MedRAX, the first versatile AI agent that seamlessly integrates state-of-the-art CXR analysis tools and multimodal large language models into a unified framework. MedRAX dynamically leverages these models to address complex medical queries without requiring additional training. To rigorously evaluate its capabilities, we introduce ChestAgentBench, a comprehensive benchmark containing 2,500 complex medical queries across 7 diverse categories. Our experiments demonstrate that MedRAX achieves state-of-the-art performance compared to both open-source and proprietary models, representing a significant step toward the practical deployment of automated CXR interpretation systems. Data and code have been publicly available at https://github.com/bowang-lab/MedRAX

For agents

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

get_harvested_code_for_paper("2502.02673")
get_code_for_paper("2502.02673")
have("2502.02673")

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

curl https://syntology.ai/api/ran/2502.02673.json

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

[![Syntology run record](https://syntology.ai/api/ran/2502.02673.svg)](https://syntology.ai/paper/2502.02673)

Connect an agent — have() is free.