SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1907.08268 · 2019

Discrete Object Generation with Reversible Inductive Construction

arXiv · PDF · Open in the Atlas

Code that ran

4 of 14 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 0 failures of our instrument; 10 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/1907.08268.json · badge: /api/ran/1907.08268.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
PrincetonLIPS/reversible-inductive-construction canonical (classifier label)
classifier basis: census
4 ran · 10 no recorded run
FunctionStatusWhere it lives
aggregate_by_incidence Ran
ran on a synthesized input
PrincetonLIPS/reversible-inductive-construction/code/genric/model/message_passing.py
code served (permissive licence) · get_code("7607b89dd113c5d9")
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
import collections

torch.manual_seed(0)

# Test case 1: incidence as torch.Tensor (sparse COO format)
values = torch.randn(2, 4)
incidence_indices = torch.tensor([[0, 1], [0, 1]])
incidence_values = torch.tensor([1.0, 1.0])
incidence_sparse = torch.sparse_coo_tensor(incidence_indices, incidence_values, (2, 2))

ARGS = {
    "values": values,
    "incidence": incidence_sparse
}
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
aggregate_by_incidence_impl Ran
ran on a synthesized input
PrincetonLIPS/reversible-inductive-construction/code/genric/model/message_passing.py
code served (permissive licence) · get_code("70f6038f3ec324cb")
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
import torch.nn

torch.manual_seed(0)

# Create small tensors for the function
# values: shape [batch_size, feature_dim] - the values to aggregate
values = torch.randn(2, 4)

# incidence: shape [2, num_edges] - edge indices
# incidence[0] contains target indices, incidence[1] contains source indices
incidence = torch.tensor([
    [0, 1, 0],  # target indices
    [0, 1, 1]   # source indices
], dtype=torch.long)

# scale: optional scaling factors, shape [num_edges]
scale = torch.tensor([0.5, 1.0, 1.5])

# output_size: optional output size (if None, defaults to values.shape[0])
output_size = 2

ARGS = {
    "values": values,
    "incidence": incidence,
    "scale": scale,
    "output_size": output_size
}
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
minmax Ran
ran on a synthesized input
PrincetonLIPS/reversible-inductive-construction/code/genric/laman/joint_network.py
code served (permissive licence) · get_code("7f42965b0968e4e8")
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, ndim 3, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create two small tensors with distinct shapes for testing
a = torch.randn(2, 4)
b = torch.randn(2, 4)

ARGS = {
    "a": a,
    "b": b,
}
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, ndim 3, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
multi_classification_prediction Ran
ran on a synthesized input
PrincetonLIPS/reversible-inductive-construction/code/genric/model/classification.py
code served (permissive licence) · get_code("e578849827120611")
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.nn as nn

torch.manual_seed(0)

# Create a mock multi_logit module with segment_multi_argmax function
class MockMultiLogit:
    @staticmethod
    def segment_multi_argmax(task_logits, scopes):
        # Returns (None, predictions)
        # predictions should be a tensor of argmax values per segment
        batch_size = scopes[0].shape[0]
        predictions = []
        for i in range(batch_size):
            logits = torch.cat([torch.narrow(tl, 0, s[i, 0], s[i, 1]) for tl, s in zip(task_logits, scopes)])
            predictions.append(torch.argmax(logits).int())
        return (None, torch.stack(predictions))

# Inject the mock into the function's module
import sys
mock_multi_logit = MockMultiLogit()
sys.modules['multi_logit'] = type(sys)('multi_logit')
sys.modules['multi_logit'].segment_multi_argmax = MockMultiLogit.segment_multi_argmax

# Create sample data
# logits_and_scopes is a tuple of (task_logits, scopes)
# task_logits is a list of logit tensors
# scopes is a list of scope tensors indicating [start, length] for each sample

# Create 2 task logits with different sizes
task_logits = [
    torch.randn(8),  # First task logits, 8 values total
    torch.randn(6),  # Second task logits, 6 values total
]

# Create scopes for batch size 2
# Each scope is [start_index, length] for that task
batch_size = 2
scopes = [
    torch.tensor([[0, 3], [3, 5]], dtype=torch.long),  # First task scopes
    torch.tensor([[0, 2], [2, 4]], dtype=torch.long),  # Second task scopes
]

logits_and_scopes = (task_logits, scopes)

ARGS = {
    "logits_and_scopes": logits_and_scopes,
    "predict": True,
    "num_samples": 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 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
comp_mols Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/molecule_edit.py
code served (permissive licence) · get_code("9f98b1761accdc8a")
How it ran

no invocation recorded.

get_mol Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/chemutils.py
code served (permissive licence) · get_code("7e083b66cf9e7a92")
How it ran

no invocation recorded.

get_mol_2D Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/chemutils.py
code served (permissive licence) · get_code("ba2ac5b259eca2e6")
How it ran

no invocation recorded.

get_smiles_2D Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/chemutils.py
code served (permissive licence) · get_code("37b4b9d7aba88ad0")
How it ran

no invocation recorded.

get_vocab Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/data_utils.py
code served (permissive licence) · get_code("f8f8fa3bcc697f45")
How it ran

no invocation recorded.

make_action_mpn Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/model/message_passing.py
code served (permissive licence) · get_code("2396d7af408f88fb")
How it ran

no invocation recorded.

multi_classification_coarse_to_fine_loss Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/model/classification.py
code served (permissive licence) · get_code("4672bd596d3a6281")
How it ran

no invocation recorded.

multi_classification_loss Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/model/classification.py
code served (permissive licence) · get_code("92535356e9126ff1")
How it ran

no invocation recorded.

my_explicit_valence Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/molecule_edit.py
code served (permissive licence) · get_code("d1e046b976a5631c")
How it ran

no invocation recorded.

my_implicit_valence Not yet run
unverified (harvested, no recorded run)
PrincetonLIPS/reversible-inductive-construction/code/genric/molecule_edit.py
code served (permissive licence) · get_code("60a5e5ecfd3deb69")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 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 14 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

The success of generative modeling in continuous domains has led to a surge of interest in generating discrete data such as molecules, source code, and graphs. However, construction histories for these discrete objects are typically not unique and so generative models must reason about intractably large spaces in order to learn. Additionally, structured discrete domains are often characterized by strict constraints on what constitutes a valid object and generative models must respect these requirements in order to produce useful novel samples. Here, we present a generative model for discrete objects employing a Markov chain where transitions are restricted to a set of local operations that preserve validity. Building off of generative interpretations of denoising autoencoders, the Markov chain alternates between producing 1) a sequence of corrupted objects that are valid but not from the data distribution, and 2) a learned reconstruction distribution that attempts to fix the corruptions while also preserving validity. This approach constrains the generative model to only produce valid objects, requires the learner to only discover local modifications to the objects, and avoids marginalization over an unknown and potentially large space of construction histories. We evaluate the proposed approach on two highly structured discrete domains, molecules and Laman graphs, and find that it compares favorably to alternative methods at capturing distributional statistics for a host of semantically relevant metrics.

For agents

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

get_harvested_code_for_paper("1907.08268")
get_code_for_paper("1907.08268")
have("1907.08268")

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

curl https://syntology.ai/api/ran/1907.08268.json

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

[![Syntology run record](https://syntology.ai/api/ran/1907.08268.svg)](https://syntology.ai/paper/1907.08268)

Connect an agent — have() is free.