SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2601.00581 · 2026

AceFF: A State-of-the-Art Machine Learning Potential for Small Molecules

Gianni De Fabritiis, Stephen Farr, Stefan Doerr, Antonio Mirarchi, Francesc Sabanés Zariquiey

arXiv · PDF · Open in the Atlas

Code that ran

6 of 11 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (3 of them constructed an object rather than computing a result; 2 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 0 failures of our instrument; 5 no recorded run), measured 2026-11-22; 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/2601.00581.json · badge: /api/ran/2601.00581.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
isayevlab/aimnetcentral community
Papers with Code archive
6 ran · 5 no recorded run
FunctionStatusWhere it lives
AEVSV Ran
ran on a synthesized input · constructed an object; no result was computed
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("40fafbd89bfd5c5f")
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 an instance of its own class, AEVSV: the run constructed the object and did not call it, so no result was computed. 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 torch import Tensor

torch.manual_seed(0)

# Create stand-in for the ops module since it's imported in forward()
class MockOps:
    @staticmethod
    def calc_distances(data):
        # Return mock distance matrix and vector matrix
        batch_size = 2
        num_atoms = 4
        d_ij = torch.rand(batch_size, num_atoms, num_atoms)
        r_ij = torch.randn(batch_size, num_atoms, num_atoms, 3)
        return d_ij, r_ij
    
    @staticmethod
    def cosine_cutoff(d_ij, rc):
        # Simple mock cutoff function
        return torch.clamp(1.0 - (d_ij / rc) ** 3, min=0.0)
    
    @staticmethod
    def exp_expand(d_ij, shifts, eta):
        # Expand distances over Gaussian basis
        d_expanded = d_ij.unsqueeze(-1) - shifts.view(1, 1, -1)
        return torch.exp(-eta * d_expanded ** 2)

class MockNbops:
    @staticmethod
    def mask_ij_(tensor, data, fill_value):
        # Mock masking operation
        return tensor

# Monkey patch the ops and nbops modules
import sys
ops_module = type(sys)('ops')
ops_module.calc_distances = MockOps.calc_distances
ops_module.cosine_cutoff = MockOps.cosine_cutoff
ops_module.exp_expand = MockOps.exp_expand
sys.modules['ops'] = ops_module

nbops_module = type(sys)('nbops')
nbops_module.mask_ij_ = MockNbops.mask_ij_
sys.modules['nbops'] = nbops_module

# Inject into builtins so the forward method can use them
import builtins
builtins.ops = ops_module
builtins.nbops = nbops_module

ARGS = {
    'rmin': 0.8,
    'rc_s': 5.0,
    'nshifts_s': 16,
    'eta_s': None,
    'rc_v': None,
    'nshifts_v': None,
    'eta_v': None,
    'shifts_s': None,
    'shifts_v': None,
}
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 AEVSV — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a AEVSV 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
AIMNet2Base Ran
ran on a synthesized input · constructed an object; no result was computed
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("00234209f7d6c68b")
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 an instance of its own class, AIMNet2Base: the run constructed the object and did not call it, so no result was computed. 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 torch import nn, Tensor
from typing import Any

torch.manual_seed(0)

# Mock nbops module since it's imported in the actual code
class MockNBOps:
    @staticmethod
    def set_nb_mode(data: dict[str, Tensor]) -> dict[str, Tensor]:
        return data
    
    @staticmethod
    def calc_masks(data: dict[str, Tensor]) -> dict[str, Tensor]:
        if 'nbmat' in data:
            data['mask'] = torch.ones_like(data['nbmat'], dtype=torch.bool)
        return data

# Inject mock nbops into the module namespace
import sys
nbops = MockNBOps()
sys.modules['nbops'] = type(sys)('nbops')
sys.modules['nbops'].set_nb_mode = nbops.set_nb_mode
sys.modules['nbops'].calc_masks = nbops.calc_masks

# Create sample data for prepare_input
sample_data = {
    'coord': torch.randn(2, 3, dtype=torch.float32),  # 2 atoms, 3 coordinates
    'numbers': torch.tensor([[6, 8], [1, 6]], dtype=torch.int64),  # atomic numbers
    'charge': torch.tensor([0.0, 1.0], dtype=torch.float32),  # 2 molecules
    'mult': torch.tensor([1, 2], dtype=torch.float32),  # multiplicity for 2 molecules
    'nbmat': torch.randint(0, 2, (2, 4, 4), dtype=torch.int32),  # neighbor matrix
}

ARGS = {}
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 AIMNet2Base — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a AIMNet2Base 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
Embedding Ran
ran on a synthesized input · constructed an object; no result was computed
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("c87249f286c324ed")
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 an instance of its own class, Embedding: the run constructed the object and did not call it, so no result was computed. 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 torch import nn

torch.manual_seed(0)

# The Embedding class extends nn.Embedding, which requires:
# - num_embeddings: int
# - embedding_dim: int
# - padding_idx: optional int
# - init: optional dict mapping indices to tensors

# Create small tensors for initialization
init_tensor_0 = torch.tensor([1.0, 2.0, 3.0, 4.0])
init_tensor_1 = torch.tensor([5.0, 6.0, 7.0, 8.0])

init_dict = {
    0: init_tensor_0,
    1: init_tensor_1,
}

ARGS = {
    'num_embeddings': 5,
    'embedding_dim': 4,
    'padding_idx': 2,
    'init': init_dict,
}
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 Embedding — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Embedding 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
ModelMetadata Ran
ran on a synthesized input
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("97a9c54de7b12363")
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 numpy as np

torch.manual_seed(0)

ARGS = {
    "format_version": 1,
    "cutoff": 10.0,
    "needs_coulomb": True,
    "needs_dispersion": False,
    "coulomb_mode": "ewald",
    "coulomb_sr_rc": 5.0,
    "coulomb_sr_envelope": "erfc",
    "d3_params": {"damping": 4.0},
    "has_embedded_lr": True,
    "implemented_species": [1, 6, 7, 8],
    "family": "model_family_v1",
    "supports_charged_systems": True,
    "has_embedded_d3ts": False,
}
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 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
_init_ahg Ran
ran, honoured the contract we drafted
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("e15e37f465dbd1ce")
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, 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 math

torch.manual_seed(0)

ARGS = {
    "b": 2,
    "m": 4,
    "n": 3,
}
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, 4, 3) — shape (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
_init_ahg_one Ran
ran, honoured the contract we drafted
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("fbf8a7452b1deb71")
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 (5, 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 math

torch.manual_seed(0)

ARGS = {
    "m": 5,
    "n": 3
}
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 (5, 3) — shape (5, 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
AIMNet2 Not yet run
unverified (harvested, no recorded run)
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("c74d7d12cd6e2de4")
How it ran

no invocation recorded.

ConvSV Not yet run
unverified (harvested, no recorded run)
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("3301b544dabcf9f2")
How it ran

no invocation recorded.

MLP Not yet run
unverified (harvested, no recorded run)
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("7e43e03f34753829")
How it ran

no invocation recorded.

get_init_module Not yet run
unverified (harvested, no recorded run)
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("8c943a3bb3a03d47")
How it ran

no invocation recorded.

get_module Not yet run
unverified (harvested, no recorded run)
isayevlab/aimnetcentral/aimnet/models/aimnet2.py
code served (permissive licence) · get_code("9cdb3130232197c6")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 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 11 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

We introduce AceFF, a pre-trained machine learning interatomic potential (MLIP) optimized for small molecule drug discovery. While MLIPs have emerged as efficient alternatives to Density Functional Theory (DFT), generalizability across diverse chemical spaces remains difficult. AceFF addresses this via a refined TensorNet2 architecture trained on a comprehensive dataset of drug-like compounds. This approach yields a force field that balances high-throughput inference speed with DFT-level accuracy. AceFF fully supports the essential medicinal chemistry elements (H, B, C, N, O, F, Si, P, S, Cl, Br, I) and is explicitly trained to handle charged states. Validation against rigorous benchmarks, including complex torsional energy scans, molecular dynamics trajectories, batched minimizations, and tests of force and energy accuracy, demonstrates that AceFF is state-of-the-art for organic molecules in the accuracy and speed 1

For agents

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

get_harvested_code_for_paper("2601.00581")
get_code_for_paper("2601.00581")
have("2601.00581")

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

curl https://syntology.ai/api/ran/2601.00581.json

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

[![Syntology run record](https://syntology.ai/api/ran/2601.00581.svg)](https://syntology.ai/paper/2601.00581)

Connect an agent — have() is free.