SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2410.22388 · NeurIPS · 2024

ET-Flow: Equivariant Flow-Matching for Molecular Conformer Generation

Dominique Beaini, Nikhil Shenoy, Hannes Stärk, Majdi Hassan, Jungyoon Lee, Stephan Thaler

arXiv · PDF · Open in the Atlas

Code that ran

12 of 12 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; 4 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 8 failures of our instrument; 0 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/2410.22388.json · badge: /api/ran/2410.22388.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
shenoynikhil/ETFlow canonical (classifier label)
classifier basis: census
12 ran · 0 no recorded run
FunctionStatusWhere it lives
GetNumRings Ran
ran, honoured the contract we drafted
shenoynikhil/ETFlow/etflow/commons/utils.py
code served (permissive licence) · get_code("401a46bdeb5e828e")
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

class MockAtom:
    """Stand-in atom object that supports IsInRingSize method"""
    def __init__(self, ring_sizes=None):
        # ring_sizes is a set of ring sizes this atom is in
        if ring_sizes is None:
            ring_sizes = {3, 5}
        self.ring_sizes = ring_sizes
    
    def IsInRingSize(self, size):
        """Check if atom is in a ring of given size"""
        return size in self.ring_sizes

# Create a mock atom that is in rings of size 3 and 5
atom = MockAtom(ring_sizes={3, 5})

ARGS = {
    "atom": atom
}
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
batched_sampling Ran
ran, honoured the contract we drafted
shenoynikhil/ETFlow/etflow/commons/sample.py
code served (permissive licence) · get_code("c1cde0272d55da1c")
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 ndarray, ndim 3, dtype family float, shape (4, 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
import torch.nn as nn
import numpy as np
from torch_geometric.data import Data

torch.manual_seed(0)

# Create a minimal model that implements the sample method
class MinimalModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def sample(self, z, bond_index, batch, node_attr, n_timesteps, 
               chiral_index, chiral_nbr_index, chiral_tag, s_churn, 
               t_min, t_max, std, sampler_type):
        # Return dummy positions: batch_size * num_nodes, 3
        # z shape: (batch_size * num_nodes,)
        num_atoms = z.shape[0]
        positions = torch.randn(num_atoms, 3, device=z.device)
        return positions

# Create a minimal Data object with all required attributes
data = Data(
    atomic_numbers=torch.tensor([6, 8, 6, 7], dtype=torch.long),  # 4 atoms
    edge_index=torch.tensor([[0, 1, 1, 2], [1, 0, 2, 1]], dtype=torch.long),  # 4 edges
    node_attr=torch.randn(4, 4),  # 4 atoms, 4 features
    batch=torch.tensor([0, 0, 0, 0], dtype=torch.long),  # all in batch 0
    chiral_index=torch.tensor([0, 2], dtype=torch.long),  # 2 chiral atoms
    chiral_nbr_index=torch.tensor([[1, 3], [1, 3]], dtype=torch.long),  # neighbors
    chiral_tag=torch.tensor([1, -1], dtype=torch.long),  # chiral tags
)

model = MinimalModel()

ARGS = {
    'model': model,
    'data': data,
    'max_batch_size': 2,
    'num_samples': 4,
    'n_timesteps': 50,
    'seed': 42,
    'device': 'cpu',
    's_churn': 1.0,
    't_min': 1.0,
    't_max': 1.0,
    'std': 1.0,
    'sampler_type': 'ode'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pytorch_lightning, torch, torch_geometric — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (4, 4, 3) — shape (4, 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
cache_decorator Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/commons/featurization.py
code served (permissive licence) · get_code("b3df40de2a0c40f5")
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
from typing import Callable

torch.manual_seed(0)

def cache_decorator(func: Callable):
    """Decorator to handle caching logic."""

    def wrapper(self, smiles: str, *args, **kwargs):
        cache_key = func.__name__
        if smiles in self.cache and cache_key in self.cache[smiles]:
            return self.cache[smiles][cache_key]
        result = func(self, smiles, *args, **kwargs)
        self.cache[smiles][cache_key] = result
        return result
    return wrapper


class MockMoleculeProcessor:
    """Mock class with cache attribute to test the decorator."""
    def __init__(self):
        self.cache = {}
    
    @cache_decorator
    def process_molecule(self, smiles: str):
        """Sample function to be decorated."""
        return {"smiles": smiles, "features": torch.randn(4, 3)}


def sample_func(obj):
    """Function that uses cache_decorator."""
    smiles = "CCO"
    result = obj.process_molecule(smiles)
    return result


mock_processor = MockMoleculeProcessor()

ARGS = {
    "func": sample_func
}
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
center_of_mass Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/models/utils.py
code served (permissive licence) · get_code("d0b50e7543c09500")
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 (6, 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)

# Construct arguments for center_of_mass function
# x: tensor of node features (num_nodes, feature_dim)
# dim: dimension to reduce over (default 0)
# batch: batch assignment for each node (default None)

x = torch.randn(6, 4)  # 6 nodes, 4 feature dimensions
dim = 0
batch = torch.tensor([0, 0, 0, 1, 1, 1], dtype=torch.long)  # 2 graphs with 3 nodes each

ARGS = {
    'x': x,
    'dim': dim,
    '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 Tensor, ndim 2, dtype family float, shape (6, 4) — shape (6, 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
center_pos Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/models/utils.py
code served (permissive licence) · get_code("405da823363ef4cc")
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 (6, 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 as nn

torch.manual_seed(0)

# pos: a tensor of shape (num_nodes, num_features)
# batch: a tensor of shape (num_nodes,) indicating which batch each node belongs to
pos = torch.randn(6, 4)  # 6 nodes, 4 features
batch = torch.tensor([0, 0, 0, 1, 1, 1])  # 2 batches, 3 nodes each

ARGS = {
    "pos": pos,
    "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 Tensor, ndim 2, dtype family float, shape (6, 4) — shape (6, 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
correct_tensor_shape Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/models/loss.py
code served (permissive licence) · get_code("4763472782ad1162")
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, 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)

# The function takes a torch.Tensor and checks its dimensionality
# If 1D, it adds a dimension via unsqueeze(1)
# If not 1D, it returns as-is
# We'll test with a 1D tensor to exercise the main branch

t_1d = torch.randn(4)

ARGS = {
    "t": t_1d
}
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 (4, 1) — shape (4, 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
l1_loss Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/models/loss.py
code served (permissive licence) · get_code("31c2e75982e527f8")
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 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 torch

torch.manual_seed(0)

# Create two tensors with distinct shapes for testing
# prediction: batch_size=2, features=4
prediction = torch.randn(2, 4)

# target: batch_size=2, features=4
target = torch.randn(2, 4)

ARGS = {
    "prediction": prediction,
    "target": target,
}
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 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
linear_schedule Ran
ran, honoured the contract we drafted
shenoynikhil/ETFlow/etflow/models/utils.py
code served (permissive licence) · get_code("21f5c6553e54587f")
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
from torch.nn.functional import pad

torch.manual_seed(0)

ARGS = {
    'low': 0.1,
    'high': 1.0,
    'max_steps': 5,
    'total_steps': 10
}
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
load_json Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/commons/io.py
code served (permissive licence) · get_code("0d62ee150d99b901")
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 sample data
temp_dir = tempfile.mkdtemp()
json_path = os.path.join(temp_dir, "test.json")

sample_data = {
    "key1": "value1",
    "key2": [1, 2, 3],
    "key3": {"nested": "data"}
}

with open(json_path, 'w') as f:
    json.dump(sample_data, f)

ARGS = {
    "path": 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 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_memmap Ran
ran, honoured the contract we drafted
shenoynikhil/ETFlow/etflow/commons/io.py
code served (permissive licence) · get_code("7645fe1b0302a652")
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 memmap. 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 numpy as np
import tempfile
import os

# Create a temporary directory and file for the memmap
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, 'test_memmap.dat')

# Create a simple memmap file with some data
data = np.arange(24, dtype=np.float32).reshape(6, 4)
memmap_write = np.memmap(temp_file, dtype=np.float32, mode='w+', shape=(6, 4))
memmap_write[:] = data
memmap_write.flush()
del memmap_write

ARGS = {
    'path': temp_file,
    'dtype': np.float32
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type memmap — a memmap output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a memmap output the recorded ndim 1 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_pkl Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/commons/io.py
code served (permissive licence) · get_code("1130080852cd8e61")
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 pickle
import tempfile

# Create a temporary directory and file
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test_data.pkl")

# Create a valid pickle file with some test data
test_data = {"key": "value", "number": 42, "list": [1, 2, 3]}
with open(temp_file, 'wb') as f:
    pickle.dump(test_data, f)

ARGS = {
    "file_path": temp_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
mse_loss Ran
ran; our contract draft was wrong, not the code
shenoynikhil/ETFlow/etflow/models/loss.py
code served (permissive licence) · get_code("0a1255a314f3751e")
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 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 torch

torch.manual_seed(0)

# Create small tensors for prediction and target
# prediction and target can be either 1D or 2D tensors
# The function will unsqueeze 1D tensors to 2D
prediction = torch.randn(2, 4)  # batch_size=2, features=4
target = torch.randn(2, 4)      # batch_size=2, features=4

ARGS = {
    "prediction": prediction,
    "target": target,
}
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 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

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 12 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

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

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

Repositories linked to this paper

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

Abstract

Predicting low-energy molecular conformations given a molecular graph is an important but challenging task in computational drug discovery. Existing stateof-the-art approaches either resort to large scale transformer-based models that diffuse over conformer fields, or use computationally expensive methods to generate initial structures and diffuse over torsion angles. In this work, we introduce Equivariant Transformer Flow (ET-Flow). We showcase that a well-designed flow matching approach with equivariance and harmonic prior alleviates the need for complex internal geometry calculations and large architectures, contrary to the prevailing methods in the field. Our approach results in a straightforward and scalable method that directly operates on all-atom coordinates with minimal assumptions. With the advantages of equivariance and flow matching, ET-Flow significantly increases the precision and physical validity of the generated conformers, while being a lighter model and faster at inference. Code is available https://github.com/shenoynikhil/ETFlow.

For agents

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

get_harvested_code_for_paper("2410.22388")
get_code_for_paper("2410.22388")
have("2410.22388")

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

curl https://syntology.ai/api/ran/2410.22388.json

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

[![Syntology run record](https://syntology.ai/api/ran/2410.22388.svg)](https://syntology.ai/paper/2410.22388)

Connect an agent — have() is free.