SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2404.11199 · 2024

RiboDiffusion: Tertiary Structure-based RNA Inverse Folding with Generative Diffusion Models

arXiv · PDF · Open in the Atlas

Code that ran

19 of 21 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, 0 violated it and 1 has no contract verdict; 16 failures of our instrument; 2 no recorded run); 1 of them was also run on the host, not in the jail, by the peer-agreement lane that set its level, 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/2404.11199.json · badge: /api/ran/2404.11199.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
ml4bio/ribodiffusion canonical (classifier label)
classifier basis: census
19 ran · 2 no recorded run
FunctionStatusWhere it lives
interpolate_fn Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/diffusion/noise_schedule.py
code served (permissive licence) · get_code("6a35b62fbc80f70a")
How it ran

Ran in Syntology's jail on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 2, dtype family float, shape (2, 1) as a later run recorded it. The image it ran in was not recorded. The comparison behind this sample's level is a different run: the peer-agreement lane ran it again on the host, not in the jail (report dated 2026-09-05T00:15:28Z). 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)

# x: PyTorch tensor with shape [N, C], where N is batch size, C is number of channels
x = torch.randn(2, 1)

# xp: PyTorch tensor with shape [C, K], where K is number of keypoints
xp = torch.linspace(0, 1, 4).unsqueeze(0).repeat(1, 1)  # shape [1, 4]

# yp: PyTorch tensor with shape [C, K]
yp = torch.linspace(0, 2, 4).unsqueeze(0).repeat(1, 1)  # shape [1, 4]

ARGS = {
    "x": x,
    "xp": xp,
    "yp": yp,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 1) — shape (2, 1) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
expand_dims Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/diffusion/noise_schedule.py
code served (permissive licence) · get_code("e6110366588c5c65")
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 Tensor, ndim 4, dtype family float, shape (2, 1, 1, 1) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a small 1D tensor with shape [N]
v = torch.randn(2)

# Create a dims parameter - total number of dimensions desired
dims = 4

ARGS = {
    "v": v,
    "dims": dims,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 4, dtype family float, shape (2, 1, 1, 1) — shape (2, 1, 1, 1) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
expand_dims Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/sampling.py
code served (permissive licence) · get_code("ce3c83921e12db4b")
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, 4, 1, 1) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Analyze the function:
# expand_dims(v, dims) takes a tensor/array v and an integer dims
# It returns v[(...,) + (None,) * (dims - 1)]
# This adds (dims - 1) new axes at the end

# Create a small test tensor
v = torch.randn(2, 4)

# dims should be an integer specifying how many dimensions to add
# If dims=3, it adds 2 new dimensions (dims - 1)
dims = 3

ARGS = {
    "v": v,
    "dims": dims
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float, shape (2, 4, 1, 1) — shape (2, 4, 1, 1) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
geo_batch Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/models/GVP_diff.py
code served (permissive licence) · get_code("bdd8f1d250ba6e52")
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_geometric
from torch_geometric.data import Batch

torch.manual_seed(0)

# The function expects batch to be a dict with keys:
# 'z_t', 'seq', 'coords', 'node_s', 'node_v', 'edge_s', 'edge_v', 'edge_index', 'mask'
# Each should have batch_size as first dimension and length as second dimension (for z_t at least)

batch_size = 2
length = 4
num_nodes = 3
num_edges = 5

batch = {
    'z_t': torch.randn(batch_size, length, 4),  # (2, 4, 4)
    'seq': torch.randint(0, 20, (batch_size, length)),  # (2, 4)
    'coords': torch.randn(batch_size, length, 3),  # (2, 4, 3)
    'node_s': torch.randn(batch_size, num_nodes, 4),  # (2, 3, 4)
    'node_v': torch.randn(batch_size, num_nodes, 3),  # (2, 3, 3)
    'edge_s': torch.randn(batch_size, num_edges, 4),  # (2, 5, 4)
    'edge_v': torch.randn(batch_size, num_edges, 3),  # (2, 5, 3)
    'edge_index': torch.randint(0, num_nodes, (batch_size, 2, num_edges)),  # (2, 2, 5)
    'mask': torch.ones(batch_size, length, dtype=torch.bool),  # (2, 4)
}

ARGS = {
    'batch': batch
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torch_geometric — 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
get_data_inverse_scaler Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/utils.py
code served (permissive licence) · get_code("faf4a5d59e3f19ad")
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
import numpy as np

class Config:
    def __init__(self):
        self.data = DataConfig()

class DataConfig:
    def __init__(self, seq_centered=True):
        self.seq_centered = seq_centered

ARGS = {
    "config": Config()
}
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 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
get_data_scaler Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/utils.py
code served (permissive licence) · get_code("f6477cfb5f1e7189")
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
import numpy as np

class Config:
    def __init__(self, seq_centered):
        self.data = type('obj', (object,), {'seq_centered': seq_centered})()

config = Config(seq_centered=True)

ARGS = {
    'config': config
}
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 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
get_orientations Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/datasets/utils.py
code served (permissive licence) · get_code("87d54062a72822b1")
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, 5, 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.nn.functional as F

torch.manual_seed(0)

# X is used with slicing operations: X[:, 1:], X[:, :-1]
# It should be a 3D tensor of shape (batch, sequence_length, features)
# Using batch=2, sequence_length=5, features=4 for distinct dimensions
X = torch.randn(2, 5, 4)

ARGS = {
    "X": X
}
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 4, dtype family float, shape (2, 5, 2, 4) — shape (2, 5, 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_orientations_single Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/datasets/utils.py
code served (permissive licence) · get_code("ecf2c47ec46adb5a")
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 (5, 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.nn.functional as F

torch.manual_seed(0)

# X should be a 2D tensor representing a sequence of positions
# The function computes orientations by taking differences between consecutive points
# Shape: (sequence_length, feature_dim)
# Using distinct sizes: 5 points, 4 features
X = torch.randn(5, 4)

ARGS = {
    "X": X
}
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 (5, 2, 4) — shape (5, 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_polynomial_schedule Ran
ran, honoured the contract we drafted
ml4bio/ribodiffusion/diffusion/noise_schedule.py
code served (permissive licence) · get_code("615f5bcf927413ba")
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 1, dtype family float, shape (10,) 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)

ARGS = {
    "time_steps": 10,
    "s": 0.0001,
    "power": 2,
}
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 1, dtype family float, shape (10,) — shape (10,) 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_posenc Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/datasets/utils.py
code served (permissive licence) · get_code("033220cc609de403")
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 (4, 16) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import torch

torch.manual_seed(0)

# Construct edge_index as a tensor with shape (2, num_edges)
# The function accesses edge_index[0] and edge_index[1], so it needs 2 rows
# Keep it small: 2 rows, 4 columns (4 edges)
edge_index = torch.tensor([
    [0, 1, 2, 3],
    [1, 2, 3, 0]
], dtype=torch.float32)

num_posenc = 16

ARGS = {
    "edge_index": edge_index,
    "num_posenc": num_posenc
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 16) — shape (4, 16) 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_rotation_frames Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ml4bio/ribodiffusion/models/esm_utils.py
code served (permissive licence) · get_code("dcd598af9c809427")
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, 4, 3, 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)

# The function get_rotation_frames expects coords of shape (batch_size x length x 3 x 3)
# where the third dimension has N, CA, C positions (3 atoms)
# and the last dimension has x, y, z coordinates (3 coordinates)

batch_size = 2
length = 4

# Create a tensor of shape (batch_size, length, 3, 3)
# This represents 2 batches, each with 4 residues, each with 3 atoms (N, CA, C),
# each with 3 coordinates (x, y, z)
coords = torch.randn(batch_size, length, 3, 3)

ARGS = {
    "coords": coords
}
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 4, dtype family float, shape (2, 4, 3, 3) — shape (2, 4, 3, 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
get_sampling_fn Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/sampling.py
code served (permissive licence) · get_code("e7cc2a0ff746b735")
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
import torch.nn as nn

torch.manual_seed(0)

# Minimal config class
class Config:
    def __init__(self):
        self.device = 'cpu'
        self.model = ModelConfig()
        self.eval = EvalConfig()

class ModelConfig:
    def __init__(self):
        self.self_cond = False
        self.pred_data = True

class EvalConfig:
    def __init__(self):
        self.cond_scale = -1.0
        self.dynamic_threshold = False
        self.dynamic_thresholding_percentile = 0.95

# Minimal noise scheduler
class NoiseScheduler:
    def __init__(self):
        self.T = 1.0
    
    def marginal_prob(self, t):
        # Return (alpha_t, sigma_t) 
        alpha_t = torch.cos(t * torch.pi / 2)
        sigma_t = torch.sin(t * torch.pi / 2)
        return alpha_t, sigma_t

# Minimal inverse scaler
class InverseScaler:
    def __call__(self, x):
        # Normalize to [0, 1] range
        return torch.sigmoid(x)

# Minimal model
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 8)
    
    def forward(self, batch, time=None, noise_level=None, cond_x=None, cond_drop_prob=0.0):
        # Return logits with shape (batch_size, seq_len, vocab_size)
        z_t = batch['z_t']
        # z_t shape: (2, 5, 4) -> output (2, 5, 8)
        out = self.linear(z_t)
        return out
    
    def eval(self):
        super().eval()
        return self

# Build ARGS
config = Config()
noise_scheduler = NoiseScheduler()
sampling_steps = 3
inverse_scaler = InverseScaler()

ARGS = {
    'config': config,
    'noise_scheduler': noise_scheduler,
    'sampling_steps': sampling_steps,
    'inverse_scaler': inverse_scaler,
    'eps': 0.001
}
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 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
modulate Ran
ran, honoured the contract we drafted
ml4bio/ribodiffusion/models/transformer_layer.py
code served (permissive licence) · get_code("62fcb3912a967a50")
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 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 small tensors for the modulate function
x = torch.randn(2, 4)  # batch size 2, dimension 4
shift = torch.randn(2, 4)  # same shape as x
scale = torch.randn(2, 4)  # same shape as x

ARGS = {
    "x": x,
    "shift": shift,
    "scale": scale,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
normalize Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/models/esm_utils.py
code served (permissive licence) · get_code("cf1378bf2f1c055e")
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 a small tensor with some NaN values and normal values
# Shape: (2, 4) - batch size 2, dimension 4
tensor = torch.randn(2, 4)
tensor[0, 1] = float('nan')  # Add a NaN value
tensor[1, 2] = float('inf')  # Add an inf value to test isfinite

ARGS = {
    "tensor": tensor,
    "dim": -1,
}
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
post_process Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/sampling.py
code served (permissive licence) · get_code("f06cabb9ba084d9c")
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 int, 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
import torch.nn as nn

torch.manual_seed(0)

# Create a small inverse_scaler callable that takes a tensor and returns a tensor
# The function calls inverse_scaler(gen_seq) and expects output suitable for argmax
class InverseScaler:
    def __call__(self, x):
        # Return a tensor with same shape as input, with values in a reasonable range
        # Since argmax will be called on dim=-1, we need the last dimension to have meaningful values
        return x * 2.0 + 1.0

inverse_scaler = InverseScaler()

# Create gen_seq: a small tensor
# Shape: (batch=2, sequence_length=3, num_classes=4)
# This allows argmax(dim=-1) to work, returning shape (2, 3)
gen_seq = torch.randn(2, 3, 4)

ARGS = {
    "gen_seq": gen_seq,
    "inverse_scaler": inverse_scaler,
}
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 int, 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
register_model Ran
ran on a synthesized input
ml4bio/ribodiffusion/models/utils.py
code served (permissive licence) · get_code("0a25a428e86e4faa")
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 type. 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

# Initialize the models registry
_MODELS = {}

# Define a minimal model class for testing
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Test case 1: Using decorator without arguments (cls is passed directly)
class TestModel1(nn.Module):
    def __init__(self):
        super().__init__()
        self.layer = nn.Linear(2, 1)

# Test case 2: Using decorator with name argument
class TestModel2(nn.Module):
    def __init__(self):
        super().__init__()
        self.layer = nn.Linear(3, 2)

# Prepare arguments for both calling patterns
# Pattern 1: register_model(cls) - direct class decoration
ARGS_DIRECT = {
    'cls': TestModel1,
    'name': None
}

# Pattern 2: register_model(name='custom_name') - returns decorator
ARGS_WITH_NAME = {
    'cls': None,
    'name': 'CustomModelName'
}

# For execution, we'll use the direct pattern as the primary test
ARGS = ARGS_DIRECT
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 type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
restore_checkpoint Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/utils.py
code served (permissive licence) · get_code("0a77b5d9d6bc5946")
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 torch
import torch.nn as nn
import tempfile
import os

torch.manual_seed(0)

# Create a simple model
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create a simple optimizer
model = SimpleModel()
optimizer = torch.optim.Adam(model.parameters(), lr=0.001)

# Create a simple EMA (Exponential Moving Average) module
class SimpleEMA(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

ema = SimpleEMA()

# Create a state dict
state = {
    'model': model,
    'optimizer': optimizer,
    'ema': ema,
    'step': 0
}

# Create a temporary checkpoint file that doesn't exist
ckpt_dir = os.path.join(tempfile.gettempdir(), 'nonexistent_checkpoint_dir', 'model.pt')

device = 'cpu'

ARGS = {
    'ckpt_dir': ckpt_dir,
    'state': state,
    'device': device
}
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 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
rotate Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/models/esm_utils.py
code served (permissive licence) · get_code("e17f40e6733c98b4")
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, 2, 4, 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 a vector tensor of shape (length=2, batch_size=2, channels=4, 3)
v = torch.randn(2, 2, 4, 3)

# Create a rotation matrix tensor of shape (length=2, batch_size=2, 3, 3)
R = torch.randn(2, 2, 3, 3)

ARGS = {
    "v": v,
    "R": R,
}
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 4, dtype family float, shape (2, 2, 4, 3) — shape (2, 2, 4, 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
tuple_index Ran
ran; our contract draft was wrong, not the code
ml4bio/ribodiffusion/models/GVP_diff.py
code served (permissive licence) · get_code("de724c9930a6f2be")
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)

# Create a tuple (s, V) where:
# s is a tensor of shape (batch_size, ...) 
# V is a tensor of shape (batch_size, ...)
# We'll use batch_size=2 with distinct dimensions

s = torch.randn(2, 4)  # shape (2, 4)
V = torch.randn(2, 3)  # shape (2, 3) - distinct from s to probe dimensions

x = (s, V)

# idx can be any object that indexes into a torch.Tensor
# Common cases: integer, slice, boolean mask, integer tensor
idx = 0  # Index into first element of batch

ARGS = {
    "x": x,
    "idx": 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 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
create_model Not yet run
unverified (harvested, no recorded run)
ml4bio/ribodiffusion/models/utils.py
code served (permissive licence) · get_code("abfc536f6d116abc")
How it ran

no invocation recorded.

get_optimizer Not yet run
unverified (harvested, no recorded run)
ml4bio/ribodiffusion/run_lib.py
code served (permissive licence) · get_code("229a69f4d6c8a2dc")
How it ran

no invocation recorded.

Also present in other papers' code

Of 21 samples lifted from this paper's linked repositories, 21 carry a function identity and 0 do not. Of the 20 distinct functions with an identity, 20 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 20 later) and 0 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
expand_dims 39 other papers
19 dated earlier, 17 later, 3 same year, order not determinable · 4 only as an assembly brick
2112.10752 (2021-12)
High-Resolution Image Synthesis with Latent Diffusion Models
2202.09778 (2022-02)
Pseudo Numerical Methods for Diffusion Models on Manifolds
2206.00927 (2022-06)
DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps
and 36 more not named here
interpolate_fn 34 other papers
16 dated earlier, 15 later, 3 same year, order not determinable
2112.10752 (2021-12)
High-Resolution Image Synthesis with Latent Diffusion Models
2211.01095 (2022-11)
DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models
2212.09478 (2022-12)
MM-Diffusion: Learning Multi-Modal Diffusion Models for Joint Audio and Video Generation
and 31 more not named here
modulate 27 other papers
1 dated earlier, 25 later, 1 undated · 2 only as an assembly brick
2312.15736 (2023-12)
Towards Real-World Blind Face Restoration with Generative Diffusion Prior
2405.17527 (2024-05)
Unisolver: PDE-Conditional Transformers Towards Universal Neural PDE Solvers
2406.11838 (2024-06)
Autoregressive Image Generation without Vector Quantization
and 24 more not named here
tuple_index 11 other papers
2 dated earlier, 9 later
2203.14486 (2022-03)
Equivariant Point Cloud Analysis via Learning Orientations for Message Passing
2309.16519 (2023-09)
AtomSurf : Surface Representation for Learning on Protein Structures
2408.11363 (2024-08)
ProteinGPT: Multimodal LLM for Protein Property Prediction and Structure Understanding
and 8 more not named here
get_optimizer 2 other papers
2 later
2410.24012 (2024-10)
Diffusion Twigs with Loop Guidance for Conditional Graph Generation
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_polynomial_schedule 2 other papers
2 later
2410.24012 (2024-10)
Diffusion Twigs with Loop Guidance for Conditional Graph Generation
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
create_model 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
geo_batch 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_data_inverse_scaler 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_data_scaler 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_orientations 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_orientations_single 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_posenc 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_rotation_frames 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
get_sampling_fn 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
normalize 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
post_process 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
register_model 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
restore_checkpoint 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model
rotate 1 other paper
1 later
2601.19232 (2026-01)
Structure-based RNA Design by Step-wise Optimization of Latent Diffusion Model

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

RNA design shows growing applications in synthetic biology and therapeutics, driven by the crucial role of RNA in various biological processes. A fundamental challenge is to find functional RNA sequences that satisfy given structural constraints, known as the inverse folding problem. Computational approaches have emerged to address this problem based on secondary structures. However, designing RNA sequences directly from 3D structures is still challenging, due to the scarcity of data, the non-unique structure-sequence mapping, and the flexibility of RNA conformation. In this study, we propose RiboDiffusion, a generative diffusion model for RNA inverse folding that can learn the conditional distribution of RNA sequences given 3D backbone structures. Our model consists of a graph neural network-based structure module and a Transformer-based sequence module, which iteratively transforms random sequences into desired sequences. By tuning the sampling weight, our model allows for a trade-off between sequence recovery and diversity to explore more candidates. We split test sets based on RNA clustering with different cut-offs for sequence or structure similarity. Our model outperforms baselines in sequence recovery, with an average relative improvement of $11\%$ for sequence similarity splits and $16\%$ for structure similarity splits. Moreover, RiboDiffusion performs consistently well across various RNA length categories and RNA types. We also apply in-silico folding to validate whether the generated sequences can fold into the given 3D RNA backbones. Our method could be a powerful tool for RNA design that explores the vast sequence space and finds novel solutions to 3D structural constraints.

For agents

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

get_harvested_code_for_paper("2404.11199")
get_code_for_paper("2404.11199")
have("2404.11199")

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

curl https://syntology.ai/api/ran/2404.11199.json

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

[![Syntology run record](https://syntology.ai/api/ran/2404.11199.svg)](https://syntology.ai/paper/2404.11199)

Connect an agent — have() is free.