SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2006.02575 · ICML · 2020

Debiased Sinkhorn barycenters

Marco Cuturi, Hicham Janati, Alexandre Gramfort

arXiv · PDF · Open in the Atlas

Code that ran

3 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 2 have no contract verdict; 1 failure of our instrument; 5 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/2006.02575.json · badge: /api/ran/2006.02575.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
hichamjanati/debiased-ot-barycenters canonical (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
google-research/ott community (classifier label)
classifier basis not recorded on the link
2 ran · 5 no recorded run
FunctionStatusWhere it lives
make_kl_loss Ran
ran on a synthesized input
google-research/ott/ott/core/quad_problems.py
code served (permissive licence) · get_code("7a03ca79356322a3")
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 2, 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 jax
import jax.numpy as jnp

ARGS = {
    "clipping_value": 1e-08,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 2, 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
register_pytree_node Ran
ran on a synthesized input
google-research/ott/ott/core/dataclasses.py
code served (permissive licence) · get_code("1b93b37ebec32b61")
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 type. 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 dataclasses
import jax

@dataclasses.dataclass
class SimpleClass:
    x: int
    y: float

ARGS = {
    "cls": SimpleClass
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
rescale_data Ran
ran; our contract draft was wrong, not the code
hichamjanati/debiased-ot-barycenters/run_barycenter_3d.py
code served (permissive licence) · get_code("8d49b2a5dc89bf80")
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) and returned SimpleMesh. 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

class SimpleMesh:
    def __init__(self):
        # Create a simple mesh with points
        self.points = np.array([
            [0.0, 0.0, 0.0],
            [1.0, 1.0, 1.0],
            [2.0, 0.5, 1.5],
            [0.5, 2.0, 0.5]
        ], dtype=np.float32)
        self.center = np.array([1.0, 1.0, 1.0])

mesh = SimpleMesh()
scale = 2.0

ARGS = {
    "mesh": mesh,
    "scale": scale
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SimpleMesh — a SimpleMesh output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleMesh 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
compute_reg_ot_cost Not yet run
unverified (harvested, no recorded run)
google-research/ott/ott/core/sinkhorn_lr.py
code served (permissive licence) · get_code("60ae0a4a27a6809c")
How it ran

no invocation recorded.

fixpoint_iter Not yet run
unverified (harvested, no recorded run)
google-research/ott/ott/core/fixed_point_loop.py
code served (permissive licence) · get_code("3f76890b8b0683ac")
How it ran

no invocation recorded.

fixpoint_iter_bwd Not yet run
unverified (harvested, no recorded run)
google-research/ott/ott/core/fixed_point_loop.py
code served (permissive licence) · get_code("d0dc7bb7a1148e3d")
How it ran

no invocation recorded.

fixpoint_iter_fwd Not yet run
unverified (harvested, no recorded run)
google-research/ott/ott/core/fixed_point_loop.py
code served (permissive licence) · get_code("5646d5b09c39906d")
How it ran

no invocation recorded.

solution_error Not yet run
unverified (harvested, no recorded run)
google-research/ott/ott/core/sinkhorn_lr.py
code served (permissive licence) · get_code("6e6033c4acd05c34")
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, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 7 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
fixpoint_iter_fwd 1 other paper
1 later
2106.06345 (2021-06)
Proximal Optimal Transport Modeling of Population Dynamics

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

Entropy regularization in optimal transport (OT) has been the driver of many recent interests for Wasserstein metrics and barycenters in machine learning. It allows to keep the appealing geometrical properties of the unregularized Wasserstein distance while having a significantly lower complexity thanks to Sinkhorn's algorithm. However, entropy brings some inherent smoothing bias, resulting for example in blurred barycenters. This side effect has prompted an increasing temptation in the community to settle for a slower algorithm such as log-domain stabilized Sinkhorn which breaks the parallel structure that can be leveraged on GPUs, or even go back to unregularized OT. Here we show how this bias is tightly linked to the reference measure that defines the entropy regularizer and propose debiased Wasserstein barycenters that preserve the best of both worlds: fast Sinkhorn-like iterations without entropy smoothing. Theoretically, we prove that the entropic OT barycenter of univariate Gaussians is a Gaussian and quantify its variance bias. This result is obtained by extending the differentiability and convexity of entropic OT to sub-Gaussian measures with unbounded supports. Empirically, we illustrate the reduced blurring and the computational advantage on various applications.

For agents

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

get_harvested_code_for_paper("2006.02575")
get_code_for_paper("2006.02575")
have("2006.02575")

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

curl https://syntology.ai/api/ran/2006.02575.json

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

[![Syntology run record](https://syntology.ai/api/ran/2006.02575.svg)](https://syntology.ai/paper/2006.02575)

Connect an agent — have() is free.