SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2408.08558 · ICLR · 2025

Linear combinations of latents in generative models: subspaces and beyond

Erik Bodin, Carl Ek, Henry Moss, Alexandru Stere, Dragos Margineantu

arXiv · PDF · Open in the Atlas

Code that ran

2 of 3 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 0 have no contract verdict; 2 failures of our instrument; 1 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/2408.08558.json · badge: /api/ran/2408.08558.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
bodin-e/linear-combinations-of-latents canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
FunctionStatusWhere it lives
_handle_singularity_at_zero Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
bodin-e/linear-combinations-of-latents/lolatents/lol_torch.py
code served (permissive licence) · get_code("05f8a360b585d2c6")
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
from jaxtyping import Float

torch.manual_seed(0)

# Create a weight tensor with shape [batch_dimension=2, num_seeds=4]
w = torch.randn(2, 4)

# Create a jitter value (small positive float)
jitter = 1e-6

ARGS = {
    'w': w,
    'jitter': jitter,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
beartype, jaxtyping, 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
lol_gaussian Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
bodin-e/linear-combinations-of-latents/lolatents/lol_torch.py
code served (permissive licence) · get_code("af5f82a4ca5e8df5")
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)

# Read the function signature and body carefully:
# - w: Float[torch.Tensor, 'batch_dimension num_seeds'] - shape (batch, num_seeds)
# - X: Union[Float[torch.Tensor, 'num_seeds num_dims'], Float[torch.Tensor, 'batch_dimension num_seeds num_dims']]
#      Either 2D or 3D tensor
# - mean: Float[torch.Tensor, 'num_dims'] - shape (num_dims,), optional
# - jitter: float - small positive value

# Create tensors with distinct sizes for clarity
batch_dimension = 2
num_seeds = 3
num_dims = 4

# w: shape (batch_dimension, num_seeds) = (2, 3)
w = torch.randn(batch_dimension, num_seeds)

# X: can be either (num_seeds, num_dims) or (batch_dimension, num_seeds, num_dims)
# Let's use the 2D version (num_seeds, num_dims) = (3, 4)
X = torch.randn(num_seeds, num_dims)

# mean: shape (num_dims,) = (4,)
mean = torch.randn(num_dims)

# jitter: small positive float
jitter = 1e-10

ARGS = {
    'w': w,
    'X': X,
    'mean': mean,
    'jitter': jitter
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
beartype, jaxtyping, 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
_check_larger_than_zero Not yet run
unverified (harvested, no recorded run)
bodin-e/linear-combinations-of-latents/lolatents/lol_torch.py
code served (permissive licence) · get_code("a4a6fec1c2071f78")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 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 3 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

Sampling from generative models has become a crucial tool for applications like data synthesis and augmentation. Diffusion, Flow Matching and Continuous Normalising Flows have shown effectiveness across various modalities, and rely on latent variables for generation. For experimental design or creative applications that require more control over the generation process, it has become common to manipulate the latent variable directly. However, existing approaches for performing such manipulations (e.g. interpolation or forming low-dimensional representations) only work well in special cases or are network or data-modality specific. We propose Latent Optimal Linear combinations (LOL) as a general-purpose method to form linear combinations of latent variables that adhere to the assumptions of the generative model. As LOL is easy to implement and naturally addresses the broader task of forming any linear combinations, e.g. the construction of subspaces of the latent space, LOL dramatically simplifies the creation of expressive low-dimensional representations of high-dimensional objects.

For agents

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

get_harvested_code_for_paper("2408.08558")
get_code_for_paper("2408.08558")
have("2408.08558")

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

curl https://syntology.ai/api/ran/2408.08558.json

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

[![Syntology run record](https://syntology.ai/api/ran/2408.08558.svg)](https://syntology.ai/paper/2408.08558)

Connect an agent — have() is free.