SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2209.15162 · ICLR · 2023

Linearly Mapping from Image to Text Space

Ellie Pavlick, Jack Merullo, Louis Castricato, Carsten Eickhoff

arXiv · PDF · Open in the Atlas

Code that ran

13 of 14 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 8 have no contract verdict; 5 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/2209.15162.json · badge: /api/ran/2209.15162.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
jmerullo/limber canonical (classifier label)
classifier basis not recorded on the link
13 ran · 1 no recorded run
FunctionStatusWhere it lives
beit_feature_extractor Ran
ran on a synthesized input
jmerullo/limber/transforms.py
code served (permissive licence) · get_code("43c05e1f4efea52f")
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 function. 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
from unittest.mock import Mock, MagicMock

torch.manual_seed(0)

# Create a mock BeitFeatureExtractor that doesn't require downloading pretrained weights
class MockBeitFeatureExtractor:
    def __call__(self, x):
        # Return a dict with 'pixel_values' key containing a batch of tensors
        # The function expects to extract [0] index, so we need a list-like structure
        pixel_values = torch.randn(1, 3, 224, 224)
        return {'pixel_values': pixel_values}

# Mock the from_pretrained method
def mock_from_pretrained(encoder_name):
    return MockBeitFeatureExtractor()

# Patch BeitFeatureExtractor
import transformers
original_beit = transformers.BeitFeatureExtractor
transformers.BeitFeatureExtractor.from_pretrained = staticmethod(mock_from_pretrained)

ARGS = {
    'encoder_name': 'microsoft/beit-base-patch16-224'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
collate_fn Ran
ran on a synthesized input
jmerullo/limber/dataset.py
code served (permissive licence) · get_code("435a97d7491369dc")
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

torch.manual_seed(0)

# The function expects:
# 1. batch_data: List[Tuple[torch.Tensor, torch.Tensor]]
#    - each tuple contains (image_tensor, caption_tensor)
#    - images will be concatenated along dim 0
#    - captions will be sliced to [:, :seq_len] then concatenated
# 2. seq_len: int (default 2048)

# Create sample batch data
# Images: shape (batch_size, channels, height, width) or similar - will be concatenated
# Captions: shape (batch_size, max_caption_len) or (batch_size, some_dim, seq_len_dim)
#           - needs to support slicing [:, :seq_len]

batch_data = [
    (
        torch.randn(2, 4),  # image tensor, batch=2, channels=4
        torch.randn(2, 5, 3)  # caption tensor, batch=2, some_dim=5, seq_dim=3
    ),
    (
        torch.randn(2, 4),  # image tensor, batch=2, channels=4
        torch.randn(2, 5, 3)  # caption tensor, batch=2, some_dim=5, seq_dim=3
    ),
]

ARGS = {
    "batch_data": batch_data,
    "seq_len": 2048,
}
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
configure_param_groups Ran
ran on a synthesized input
jmerullo/limber/config_params.py
code served (permissive licence) · get_code("b93589f28f3dd515")
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 torch.nn as nn
from collections import defaultdict

torch.manual_seed(0)

# Define a minimal config class with all required attributes
class Config:
    def __init__(self):
        self.weight_decay = 0.01
        self.image_enc_lr = None
        self.use_image_embed_layernorm = False
        self.min_lr = 1e-5
        self.lr = 1e-3
        self.deepspeed_config_params = {
            'scheduler': {
                'params': {
                    'warmup_min_lr': [],
                    'warmup_max_lr': []
                }
            }
        }

# Define minimal model components
class SimpleTransformer(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 4)
        self.bias_layer = nn.Linear(4, 4)
    
    def forward(self, x):
        return self.linear(x)

class ImagePrefix(nn.Module):
    def __init__(self):
        super().__init__()
        self.enc = nn.Linear(4, 4)
        self.proj = nn.Linear(4, 4)
        self.ln = nn.LayerNorm(4)

class FullModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.image_prefix = ImagePrefix()
        self.transformer = SimpleTransformer()
        self.lm_head = nn.Linear(4, 4)
        # Optionally add class_head
        self.class_head = nn.Linear(4, 2)
    
    def forward(self, x):
        return self.transformer(x)

# Create instances
model = FullModel()
config = Config()

# Ensure all parameters require gradients
for param in model.parameters():
    param.requires_grad = True

ARGS = {
    'model': model,
    'config': config
}
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 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
full_nfresnet50 Ran
ran on a synthesized input
jmerullo/limber/image_encoders.py
code served (permissive licence) · get_code("32d95b9de0cbb2de")
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 NormFreeNet. 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

torch.manual_seed(0)

ARGS = {
    'device': 'cpu',
    'pretrained': False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
timm, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type NormFreeNet — a NormFreeNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a NormFreeNet 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_params_for_weight_decay_optimization Ran
ran on a synthesized input
jmerullo/limber/config_params.py
code served (permissive licence) · get_code("f2aaccc3c641089b")
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 torch.nn as nn

torch.manual_seed(0)

# Create a simple module with LayerNorm, Embedding, Linear layers, and biases
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.embedding = nn.Embedding(10, 4)
        self.layer_norm = nn.LayerNorm(4)
        self.linear1 = nn.Linear(4, 3, bias=True)
        self.linear2 = nn.Linear(3, 2, bias=False)
    
    def forward(self, x):
        x = self.embedding(x)
        x = self.layer_norm(x)
        x = self.linear1(x)
        x = self.linear2(x)
        return x

# Create a simple config class
class Config:
    def __init__(self, weight_decay=0.01):
        self.weight_decay = weight_decay

# Instantiate the module and config
module = SimpleModel()
config = Config(weight_decay=0.01)

ARGS = {
    'module': module,
    'config': config
}
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 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
load_config Ran
ran; our contract draft was wrong, not the code
jmerullo/limber/config.py
code served (permissive licence) · get_code("ace12fad87592626")
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 yaml
import os
from pathlib import Path
import tempfile

# Create a temporary directory and a minimal YAML config file
temp_dir = tempfile.mkdtemp()
config_dir = Path(temp_dir)

# Create a sample config file
config_content = """
model:
  name: test_model
  layers: 3
training:
  epochs: 10
  learning_rate: 0.001
"""

config_file = config_dir / "test_config.yml"
with open(config_file, 'w') as f:
    f.write(config_content)

# Define ARGS with valid arguments for load_config
ARGS = {
    'path': str(config_file),
    'config_dir': config_dir
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
yaml — 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_json Ran
ran on a synthesized input
jmerullo/limber/dataset.py
code served (permissive licence) · get_code("26c48ad88add4d84")
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 json
import tempfile
import os

# Create a temporary JSON file with valid content
temp_dir = tempfile.mkdtemp()
temp_json_file = os.path.join(temp_dir, "test.json")

with open(temp_json_file, 'w') as f:
    json.dump({"key": "value", "number": 42}, f)

ARGS = {
    "filename": temp_json_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 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
nfresnet50 Ran
ran on a synthesized input
jmerullo/limber/image_encoders.py
code served (permissive licence) · get_code("81a63795029ff25d")
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 Sequential. 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 unittest.mock import MagicMock, patch

torch.manual_seed(0)

# Mock timm.create_model to return a simple model without downloading
def mock_create_model(model_name, pretrained=False):
    """Create a mock nf_resnet50 model"""
    model = nn.Sequential(
        nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3),
        nn.BatchNorm2d(64),
        nn.ReLU(inplace=True),
        nn.MaxPool2d(kernel_size=3, stride=2, padding=1),
        nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1),
        nn.BatchNorm2d(128),
        nn.ReLU(inplace=True),
        nn.AdaptiveAvgPool2d((7, 7)),
        nn.Flatten(),
        nn.Linear(128 * 49, 1000),
    )
    return model

# Patch timm.create_model before importing/calling nfresnet50
import sys
import types
timm_module = types.ModuleType('timm')
timm_module.create_model = mock_create_model
sys.modules['timm'] = timm_module

ARGS = {
    'device': 'cpu',
    'pretrained': False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
timm, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Sequential — a Sequential 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
pad_to_size Ran
ran; our contract draft was wrong, not the code
jmerullo/limber/transforms.py
code served (permissive licence) · get_code("e82e4764f914492f")
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 numpy as np

# Create a small PIL Image for testing
# The function expects a PIL Image object with .size attribute
test_image = Image.new('RGB', (100, 150))

ARGS = {
    'x': test_image,
    'size': 256
}
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 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
pad_to_size_tensor Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jmerullo/limber/transforms.py
code served (permissive licence) · get_code("d3af6b05f019b479")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 256, 256) 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.nn.functional as F

torch.manual_seed(0)

# Create a small tensor with shape (batch, height, width)
# Using distinct sizes: batch=2, height=4, width=3
x = torch.randn(2, 4, 3)

ARGS = {
    "x": x,
    "size": 256
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 256, 256) — shape (2, 256, 256) 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
read_img_captions Ran
ran on a synthesized input
jmerullo/limber/dataset.py
code served (permissive licence) · get_code("a4d1b53e4f72d453")
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 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 json
import tempfile
import os

# Create a temporary JSONL file with test data
temp_dir = tempfile.mkdtemp()
test_jsonl_path = os.path.join(temp_dir, "test.jsonl")

# Write test data to the JSONL file
test_data = [
    {"image": "img1.jpg", "caption": "a dog running"},
    {"image": "img2.jpg", "caption": "a cat sleeping"},
    {"other": "data", "more": "N/A"},  # This should be skipped
    {"image": "img3.jpg", "caption": "a bird flying"},
]

with open(test_jsonl_path, 'w') as f:
    for item in test_data:
        # Write as list format to match item[-1] and item[-2] indexing
        if item.get("image") and item.get("caption"):
            f.write(json.dumps([item["image"], item["caption"]]) + "\n")
        else:
            f.write(json.dumps([None, "N/A"]) + "\n")

ARGS = {
    "filename": test_jsonl_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 list, ndim 2, 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
remove_tokens_after_eos Ran
ran; our contract draft was wrong, not the code
jmerullo/limber/sampling.py
code served (permissive licence) · get_code("ba246ce8c9917aba")
How it ran

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

import torch

torch.manual_seed(0)

# Create a small tensor with token IDs
# Include some regular tokens, an eos_token, image_token, and tokens after eos
tensor = torch.tensor([101, 102, 103, 104, 105, 0, 102, 103], dtype=torch.long)

# Define token IDs
eos_token = 0      # End-of-sequence token
image_token = 104  # Image token

ARGS = {
    "tensor": tensor,
    "eos_token": eos_token,
    "image_token": image_token,
}
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 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
top_k_filter Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
jmerullo/limber/sampling.py
code served (permissive licence) · get_code("aa8cc3e8f125148a")
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 logits tensor: batch_size=2, vocab_size=4
logits = torch.randn(2, 4)

# Set k to a valid value less than vocabulary size
k = 2

ARGS = {
    "logits": logits,
    "k": k,
}
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
get_tokenizer Not yet run
unverified (harvested, no recorded run)
jmerullo/limber/limber_gptj.py
code served (permissive licence) · get_code("1fe016f046db89bd")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 1 later) and 9 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
pad_to_size 2 other papers
1 dated earlier, 1 later
2112.05253 (2021-12)
MAGMA -- Multimodal Augmentation of Generative Models through Adapter-based Finetuning
2301.13823 (2023-01)
Grounding Language Models to Images for Multimodal Inputs and Outputs
load_config 1 other paper
1 dated earlier
2112.05253 (2021-12)
MAGMA -- Multimodal Augmentation of Generative Models through Adapter-based Finetuning
pad_to_size_tensor 1 other paper
1 dated earlier
2112.05253 (2021-12)
MAGMA -- Multimodal Augmentation of Generative Models through Adapter-based Finetuning
remove_tokens_after_eos 1 other paper
1 dated earlier
2112.05253 (2021-12)
MAGMA -- Multimodal Augmentation of Generative Models through Adapter-based Finetuning
top_k_filter 1 other paper
1 dated earlier
2112.05253 (2021-12)
MAGMA -- Multimodal Augmentation of Generative Models through Adapter-based Finetuning

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

The extent to which text-only language models (LMs) learn to represent features of the non-linguistic world is an open question. Prior work has shown that pretrained LMs can be taught to caption images when a vision model's parameters are optimized to encode images in the language space. We test a stronger hypothesis: that the conceptual representations learned by frozen text-only models and vision-only models are similar enough that this can be achieved with a linear map. We show that the image representations from vision models can be transferred as continuous prompts to frozen LMs by training only a single linear projection. Using these to prompt the LM achieves competitive performance on captioning and visual question answering tasks compared to models that tune both the image encoder and text decoder (such as the MAGMA model). We compare three image encoders with increasing amounts of linguistic supervision seen during pretraining: BEIT (no linguistic information), NF-ResNET (lexical category information), and CLIP (full natural language descriptions). We find that all three encoders perform equally well at transferring visual property information to the language model (e.g., whether an animal is large or small), but that image encoders pretrained with linguistic supervision more saliently encode category information (e.g., distinguishing hippo vs. elephant) and thus perform significantly better on benchmark language-and-vision tasks. Our results indicate that LMs encode conceptual information structurally similarly to vision-based models, even those that are solely trained on images. Code is available here: https://github.com/jmerullo/limber NFRN50 Tuned : my motorcycle MAGMA (Ours) : my bike MAGMA (Rel.) : a 1973 Honda CB750 NFRN Random: the car on the road BEIT: the bike.' NFRN50: a motorcycle on a motorcycle. CLIP: my motorcycle. Ground Truth: A bicycle is standing next to a bed in a room. NFRN50 Tuned : my bike with a basket. MAGMA (Ours) : my bike MAGMA (Rel.) : a vintage bike vintage bike NFRN Random: a young boy in a blue shirt and a red hat on a bicycle BEIT: the bike. NFRN50: my bicycle with a basket on the front. CLIP: my bike, which is a fixed gear bike. A man riding a wave on top of a white surfboard. NFRN50 Tuned : a surfer in the water MAGMA (Ours) : a surfer riding a wave. MAGMA (Rel.) : a surfer riding a wave NFRN Random: the view from the top of the mountain BEIT: a surfer riding a wave. NFRN50: a surfer in the water. CLIP: a surfer riding a wave. Ground Truth: A person standing on top of a beach flying a kite. NFRN50 Tuned : a man standing on the beach with a surfboard. MAGMA (Ours) : a kite on a beach. MAGMA (Rel.) : a man running on the beach NFRN Random: the new car in the garage BEIT: a man standing on a rock in the sea. NFRN50: a man surfing on the beach. CLIP: a man flying a kite on the beach. Ground Truth: A lady explains the process of milking a cow. NFRN50 Tuned : a horse and rider at the fair. MAGMA (Ours) : a woman on a stage with kids in the audience. MAGMA (Rel.) : a woman milking a cow in front of a group of children. NFRN Random: the new restaurant. BEIT: the crowd at the event. NFRN50: a crowd of people at a festival CLIP: the dairy cows. Ground Truth: A woman is taking notes in front of her laptop NFRN50 Tuned : the man who was shot and killed in the lobby of the building. MAGMA (Ours) : a student studying. MAGMA (Rel.) : a student working on a paper NFRN Random: the man who was killed in the attack. BEIT: the room where the meeting took place. NFRN50: the room where the meeting took place. CLIP: a student working on his laptop. Ground Truth: A couple of men standing on top of a field together. NFRN50 Tuned : the team before the match MAGMA (Ours) : men playing a game of sport. MAGMA (Rel.) : a men's lacrosse game at the University of New England NFRN Random: the new home of the football team BEIT: football player and football player during a training session. NFRN50: the team before the game. CLIP: the team during a practice session.

For agents

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

get_harvested_code_for_paper("2209.15162")
get_code_for_paper("2209.15162")
have("2209.15162")

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

curl https://syntology.ai/api/ran/2209.15162.json

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

[![Syntology run record](https://syntology.ai/api/ran/2209.15162.svg)](https://syntology.ai/paper/2209.15162)

Connect an agent — have() is free.