SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2604.08648 · 2026

High-dimensional inference for the γ-ray sky with differentiable programming

Siddharth Mishra-Sharma, Yuqing Wu, Tracy Slatyer, Yitian Sun

arXiv · PDF · Open in the Atlas

Code that ran

7 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; 1 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 5 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/2604.08648.json · badge: /api/ran/2604.08648.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
smsharma/fermi-prob-prog canonical (classifier label)
classifier basis: census
4 ran · 0 no recorded run
yitiansun/gce-prob-prog canonical (classifier label)
classifier basis: census
3 ran · 1 no recorded run
FunctionStatusWhere it lives
beta_schedule Ran
ran, honoured the contract we drafted
smsharma/fermi-prob-prog/src/fpp/models/svi.py
code served (permissive licence) · get_code("c2ec4830cd3b5a0e")
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 ArrayImpl, ndim 1, dtype family float, shape (5,) 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 jax.numpy as jnp

# Construct valid arguments for beta_schedule
ARGS = {
    'num_steps': 5,
    'start_T': 10.0,
    'end_T': 1.0,
    'kind': 'geometric'
}
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 ArrayImpl, ndim 1, dtype family float, shape (5,) — shape (5,) 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
dnds Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
smsharma/fermi-prob-prog/src/fpp/models/scd.py
code served (permissive licence) · get_code("83ce5ae01ef4f1c0")
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 ArrayImpl, 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 jax.numpy as jnp
import numpy as np

# The function dnds takes two parameters: s and theta
# s appears to be a scalar or array of values
# theta is a tuple of 6 parameters: (a, n1, n2, n3, sb1, sb2)

# Create small test values
s_value = jnp.array(0.5)  # scalar value to evaluate the function at

# theta is unpacked as: a, n1, n2, n3, sb1, sb2
# These appear to be model parameters for a probability distribution
theta_value = (
    1.0,      # a: amplitude/normalization
    2.5,      # n1: power law index 1
    2.0,      # n2: power law index 2
    1.5,      # n3: power law index 3
    10.0,     # sb1: break scale 1
    5.0       # sb2: break scale 2
)

ARGS = {
    "s": s_value,
    "theta": theta_value
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
dnds_1b Ran
ran, violated the contract we drafted
smsharma/fermi-prob-prog/src/fpp/models/scd.py
code served (permissive licence) · get_code("e4eb54fa12dbc47b")
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 ArrayImpl, ndim 1, dtype family float, shape (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 jax.numpy as jnp
import numpy as np

ARGS = {
    "s": jnp.array([0.5, 1.5, 2.5, 3.5]),
    "theta": (0.1, 1.5, 2.0, 2.8)
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 1, dtype family float, shape (4,) — shape (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
downsample Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yitiansun/gce-prob-prog/utils/map_utils.py
code served (permissive licence) · get_code("177e100d810173da")
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 ndarray, ndim 2, dtype family float, shape (3, 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 numpy as np

# Create a small test array with dimensions divisible by downsampling factor
arr = np.arange(36).reshape(6, 6).astype(float)

# Define downsampling factor
factor = 2

# Define mode
mode = 'sum'

ARGS = {
    'arr': arr,
    'factor': factor,
    'mode': mode
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (3, 3) — shape (3, 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
log_like_internal Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yitiansun/gce-prob-prog/likelihoods/npll_jax.py
code served (permissive licence) · get_code("4cbda7888e72b762")
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 ArrayImpl, 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 jax.numpy as jnp
import numpy as np

# Set seeds for reproducibility
np.random.seed(0)

# Define small test data
npixROI = 2  # number of pixels in ROI
k_max = 3    # maximum k value

# pt_sum_compressed: scalar or array of log probabilities
pt_sum_compressed = jnp.array(-0.5)

# data: array of shape (npixROI,) containing indices into k dimension
# Values must be in range [0, k_max]
data = jnp.array([0, 2], dtype=jnp.int32)

# x_m_ary: array of shape (npixROI, k_max+1)
# Contains the x_m values indexed by pixel and k
x_m_ary = jnp.array([
    [0.1, 0.2, 0.3, 0.4],
    [0.15, 0.25, 0.35, 0.45]
], dtype=jnp.float32)

# x_m_sum: array of shape (npixROI,) - sum of x_m values
x_m_sum = jnp.array([1.0, 1.2], dtype=jnp.float32)

ARGS = {
    "pt_sum_compressed": pt_sum_compressed,
    "data": data,
    "x_m_ary": x_m_ary,
    "x_m_sum": x_m_sum,
    "k_max": k_max,
    "npixROI": npixROI,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
log_like_np Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yitiansun/gce-prob-prog/likelihoods/npll_jax.py
code served (permissive licence) · get_code("c52a9d4e3cebdb6f")
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 ArrayImpl, 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 jax
import jax.numpy as jnp
import numpy as np

# Define small test parameters
k_max = 3
npixROI = 2
n_templates = 2
n_f_values = 4
n_s_values = 5

# theta: tuple of 6 parameters for dnds function
# (a, n1, n2, n3, sb1, sb2)
theta = (
    jnp.array([1.5, 2.0]),           # a (2 templates)
    jnp.array([2.1, 2.2]),           # n1
    jnp.array([1.9, 1.8]),           # n2
    jnp.array([3.0, 2.9]),           # n3
    jnp.array([10.0, 11.0]),         # sb1
    jnp.array([5.0, 4.5])            # sb2
)

# pt_sum_compressed: shape (npixROI,)
pt_sum_compressed = jnp.array([0.5, 0.3], dtype=jnp.float64)

# npt_compressed: shape (npixROI,)
npt_compressed = jnp.array([2.0, 3.0], dtype=jnp.float64)

# data: shape (npixROI,), integer counts
data = jnp.array([1, 2], dtype=jnp.int32)

# f_ary: shape (n_f_values,)
f_ary = jnp.logspace(-1, 1, n_f_values, dtype=jnp.float64)

# df_rho_div_f_ary: shape (n_f_values,)
df_rho_div_f_ary = jnp.ones(n_f_values, dtype=jnp.float64) * 0.1

ARGS = {
    "theta": theta,
    "pt_sum_compressed": pt_sum_compressed,
    "npt_compressed": npt_compressed,
    "data": data,
    "f_ary": f_ary,
    "df_rho_div_f_ary": df_rho_div_f_ary,
    "k_max": k_max,
    "npixROI": npixROI,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
log_trapz Ran
ran; our contract draft was wrong, not the code
smsharma/fermi-prob-prog/src/fpp/likelihoods/npll_jax_1b.py
code served (permissive licence) · get_code("cb25b0119b0fb1ac")
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 ArrayImpl, 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 jax.numpy as jnp
import numpy as np

# Create small test data for log_trapz
# logf should be an array of log-space values
# x should be monotonically increasing coordinates

# Create x values - shape (5,) for integration points
x = jnp.array([0.0, 1.0, 2.0, 3.0, 4.0])

# Create logf values - shape (2, 5) for batch of 2, 5 integration points
# These represent log of function values at each x point
logf = jnp.array([
    [-1.0, -0.5, -0.3, -0.2, -0.1],
    [-2.0, -1.5, -1.0, -0.8, -0.5]
])

ARGS = {
    "logf": logf,
    "x": x,
    "axis": -1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, 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
interp1d Not yet run
unverified (harvested, no recorded run)
yitiansun/gce-prob-prog/templates/rigid_templates.py
code served (permissive licence) · get_code("f41c4c48d9327f0a")
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, 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 8 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 motivate the use of differentiable probabilistic programming techniques in order to account for the large model-space inherent to astrophysical γ-ray analyses. Targeting the longstanding Galactic Center γ-ray Excess (GCE) puzzle, we construct differentiable forward model and likelihood that make liberal use of GPU acceleration and vectorization in order to simultaneously account for a continuum of possible spatial morphologies consistent with the GCE emission in a fully probabilistic manner. Our setup allows for efficient inference over the large model space using variational methods. Beyond application to γ-ray data, a goal of this work is to showcase how differentiable probabilistic programming can be used as a tool to enable flexible analyses of astrophysical datasets.

For agents

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

get_harvested_code_for_paper("2604.08648")
get_code_for_paper("2604.08648")
have("2604.08648")

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

curl https://syntology.ai/api/ran/2604.08648.json

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

[![Syntology run record](https://syntology.ai/api/ran/2604.08648.svg)](https://syntology.ai/paper/2604.08648)

Connect an agent — have() is free.