SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2107.00096 · 2021

Improving black-box optimization in VAE latent space using decoder uncertainty

arXiv · PDF · Open in the Atlas

Code that ran

6 of 8 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 6 have no contract verdict; 0 failures of our instrument; 2 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/2107.00096.json · badge: /api/ran/2107.00096.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
pascalnotin/uncertainty_guided_optimization community (classifier label)
classifier basis not recorded on the link
6 ran · 2 no recorded run
FunctionStatusWhere it lives
LDE Ran
ran on a synthesized input
pascalnotin/uncertainty_guided_optimization/utils/optimization_utils.py
code served (permissive licence) · get_code("20b3d08481214b2f")
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 small log probability tensors
log_ai = torch.randn(2, 4)
log_bi = torch.randn(2, 4)

ARGS = {
    "log_ai": log_ai,
    "log_bi": log_bi,
}
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
LSE Ran
ran on a synthesized input
pascalnotin/uncertainty_guided_optimization/utils/optimization_utils.py
code served (permissive licence) · get_code("c89d89cbabceb5a8")
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 (2,) 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)

# LSE expects log_ai of dim (batch_size, num_items_in_sum)
# Create a tensor with batch_size=2, num_items_in_sum=4
log_ai = torch.randn(2, 4)

ARGS = {
    "log_ai": log_ai
}
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 (2,) — shape (2,) 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
dfs Ran
ran on a synthesized input
pascalnotin/uncertainty_guided_optimization/JTVAE/fast_jtnn/mol_tree.py
code served (permissive licence) · get_code("4027a8a9d160edc2")
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.

class Node:
    def __init__(self, idx, neighbors=None):
        self.idx = idx
        self.neighbors = neighbors if neighbors is not None else []

# Create a simple tree structure
# Tree structure:
#       0
#      / \
#     1   2
#    /
#   3

node3 = Node(idx=3, neighbors=[])
node1 = Node(idx=1, neighbors=[node3])
node2 = Node(idx=2, neighbors=[])
node0 = Node(idx=0, neighbors=[node1, node2])

# Add parent references for traversal
node3.neighbors.append(node1)
node2.neighbors.append(node0)
node1.neighbors.append(node0)

ARGS = {
    "node": node0,
    "fa_idx": -1
}
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 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
have_slots Ran
ran on a synthesized input
pascalnotin/uncertainty_guided_optimization/JTVAE/fast_jtnn/jtnn_dec.py
code served (permissive licence) · get_code("da21481c18314904")
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 bool, 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
import numpy as np

# Analyzing the function:
# - fa_slots and ch_slots are iterables of tuples
# - Each tuple has 3 elements: (a, c, h) where a and c are compared for equality,
#   and h is a numeric value that can be added
# - The tuples are accessed by index and the lists are mutated with .pop()

# Create valid slot lists with tuples of (a, c, h)
# a: can be 'C' or other strings
# c: integer or comparable value
# h: numeric value

fa_slots = [
    ('C', 1, 2),
    ('D', 1, 3),
    ('E', 2, 1)
]

ch_slots = [
    ('C', 1, 3),
    ('F', 2, 2),
    ('D', 1, 4)
]

ARGS = {
    'fa_slots': fa_slots,
    'ch_slots': ch_slots
}
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 bool, 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
onek_encoding_unk Ran
ran on a synthesized input
pascalnotin/uncertainty_guided_optimization/JTVAE/fast_jtnn/jtmpn.py
code served (permissive licence) · get_code("5fe179a1fb051e63")
How it ran

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

import torch

def onek_encoding_unk(x, allowable_set):
    if x not in allowable_set:
        x = allowable_set[-1]
    return [x == s for s in allowable_set]

# Setup: x is a value that will be checked against allowable_set
# allowable_set is a list/sequence of values
# The function returns a list of booleans indicating which elements match x

x_value = "C"
allowable_set_value = ["C", "N", "O", "unknown"]

ARGS = {
    "x": x_value,
    "allowable_set": allowable_set_value
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
read_rules Ran
ran on a synthesized input
pascalnotin/uncertainty_guided_optimization/utils/rd_filters.py
code served (permissive licence) · get_code("5ccb47bdcb3ec477")
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 tempfile
import json
import os

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

rules_content = {
    "rule1": "value1",
    "rule2": "value2",
    "nested": {
        "key": "value"
    }
}

with open(rules_file, 'w') as f:
    json.dump(rules_content, f)

ARGS = {
    "rules_file_name": rules_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
atom_features Not yet run
unverified (harvested, no recorded run)
pascalnotin/uncertainty_guided_optimization/JTVAE/fast_jtnn/jtmpn.py
code served (permissive licence) · get_code("8defa56cfba45b58")
How it ran

no invocation recorded.

get_config_file Not yet run
unverified (harvested, no recorded run)
pascalnotin/uncertainty_guided_optimization/utils/rd_filters.py
code served (permissive licence) · get_code("7aee3ccc4e667f90")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 3 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 2 later) and 5 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
atom_features 2 other papers
1 dated earlier, 1 later
1812.01070 (2018-12)
Learning Multimodal Graph-to-Graph Translation for Molecular Optimization
2502.15805 (2025-02)
FragFM: Hierarchical Framework for Efficient Molecule Generation via Fragment-Level Discrete Flow Matching
dfs 1 other paper
1 later
2206.12411 (2022-06)
Sample Efficiency Matters: A Benchmark for Practical Molecular Optimization
have_slots 1 other paper
1 dated earlier
1905.12712 (2019-05)
Path-Augmented Graph Transformer Network

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

Optimization in the latent space of variational autoencoders is a promising approach to generate high-dimensional discrete objects that maximize an expensive black-box property (e.g., drug-likeness in molecular generation, function approximation with arithmetic expressions). However, existing methods lack robustness as they may decide to explore areas of the latent space for which no data was available during training and where the decoder can be unreliable, leading to the generation of unrealistic or invalid objects. We propose to leverage the epistemic uncertainty of the decoder to guide the optimization process. This is not trivial though, as a naive estimation of uncertainty in the high-dimensional and structured settings we consider would result in high estimator variance. To solve this problem, we introduce an importance sampling-based estimator that provides more robust estimates of epistemic uncertainty. Our uncertainty-guided optimization approach does not require modifications of the model architecture nor the training process. It produces samples with a better trade-off between black-box objective and validity of the generated samples, sometimes improving both simultaneously. We illustrate these advantages across several experimental settings in digit generation, arithmetic expression approximation and molecule generation for drug design.

For agents

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

get_harvested_code_for_paper("2107.00096")
get_code_for_paper("2107.00096")
have("2107.00096")

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

curl https://syntology.ai/api/ran/2107.00096.json

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

[![Syntology run record](https://syntology.ai/api/ran/2107.00096.svg)](https://syntology.ai/paper/2107.00096)

Connect an agent — have() is free.