SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1909.09849 · 2019

Multiagent Evaluation under Incomplete Information

arXiv · PDF · Open in the Atlas

Code that ran

5 of 9 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 5 have no contract verdict; 0 failures of our instrument; 4 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/1909.09849.json · badge: /api/ran/1909.09849.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
microsoft/InfoGainalpharank community (classifier label)
classifier basis not recorded on the link
5 ran · 4 no recorded run
FunctionStatusWhere it lives
h_sample_func Ran
ran on a synthesized input
microsoft/InfoGainalpharank/src/sampling_schemes/distributions/indep_normal.py
code served (permissive licence) · get_code("49cbbd6a5ec49003")
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, 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 numpy as np

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

# Define parameters based on function signature and body analysis
strat = 1  # strategy index (integer)

num_pops = 3  # number of populations

hallucinate_samples = 5  # number of hallucinated samples to add

# counts: shape (num_pops, num_strategies)
# The function accesses counts[p][strat], so it's a 2D array
num_strategies = 4
counts = np.array([[10, 20, 15, 25],
                   [12, 18, 22, 14],
                   [15, 25, 10, 20]], dtype=np.float64)

# running_total: shape (num_pops, num_strategies)
# The function accesses running_total[p][strat]
running_total = np.array([[100.0, 200.0, 150.0, 250.0],
                          [120.0, 180.0, 220.0, 140.0],
                          [150.0, 250.0, 100.0, 200.0]], dtype=np.float64)

# starting_mu: shape (num_pops, num_strategies)
starting_mu = np.array([[0.5, 0.6, 0.4, 0.7],
                        [0.55, 0.65, 0.45, 0.75],
                        [0.48, 0.58, 0.42, 0.68]], dtype=np.float64)

# starting_var: shape (num_pops, num_strategies)
starting_var = np.array([[0.1, 0.12, 0.08, 0.15],
                         [0.11, 0.13, 0.09, 0.14],
                         [0.09, 0.11, 0.07, 0.16]], dtype=np.float64)

# noise_var: scalar or array
noise_var = np.array([[0.05, 0.05, 0.05, 0.05],
                      [0.05, 0.05, 0.05, 0.05],
                      [0.05, 0.05, 0.05, 0.05]], dtype=np.float64)

# halluc_sample: shape (num_pops, num_strategies)
# The function accesses halluc_sample[p][strat]
halluc_sample = np.array([[0.52, 0.62, 0.42, 0.72],
                          [0.57, 0.67, 0.47, 0.77],
                          [0.50, 0.60, 0.44, 0.70]], dtype=np.float64)

ARGS = {
    'strat': strat,
    'running_total': running_total,
    'counts': counts,
    'num_pops': num_pops,
    'hallucinate_samples': hallucinate_samples,
    'starting_mu': starting_mu,
    'starting_var': starting_var,
    'noise_var': noise_var,
    'halluc_sample': halluc_sample,
}
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, 4) — shape (3, 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
jsd Ran
ran on a synthesized input
microsoft/InfoGainalpharank/src/acquisition/functions.py
code served (permissive licence) · get_code("7aa5cfcba664b38c")
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 float64, ndim 0, 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
import numpy as np
from scipy.spatial.distance import jensenshannon

torch.manual_seed(0)

# The jsd function takes two arguments: x and y
# It passes them to jensenshannon, which expects array-like objects
# representing probability distributions (non-negative, sum to 1)

# Create two valid probability distributions
x = np.array([0.2, 0.3, 0.3, 0.2])
y = np.array([0.1, 0.4, 0.3, 0.2])

ARGS = {
    "x": x,
    "y": y
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
kendall_partial Ran
ran on a synthesized input
microsoft/InfoGainalpharank/src/acquisition/functions.py
code served (permissive licence) · get_code("0fd4e18250362582")
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.

import numpy as np
from scipy.stats import rankdata

# The function expects two array-like objects of numbers (p1 and p2)
# and an optional bucket_decimals parameter
# Let's create simple numpy arrays for p1 and p2

p1_data = np.array([0.1, 0.5, 0.3, 0.8, 0.2])
p2_data = np.array([0.15, 0.45, 0.35, 0.75, 0.25])

ARGS = {
    "p1": p1_data,
    "p2": p2_data,
    "bucket_decimals": 2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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
l1 Ran
ran on a synthesized input
microsoft/InfoGainalpharank/src/acquisition/functions.py
code served (permissive licence) · get_code("a1156bd23c9f5eba")
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 float64, ndim 0, 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 numpy as np

# Create small numpy arrays for x and y
x = np.array([[1.0, 2.0, 3.0, 4.0],
              [5.0, 6.0, 7.0, 8.0]])

y = np.array([[1.5, 2.5, 2.5, 4.5],
              [5.5, 5.5, 7.5, 7.5]])

ARGS = {
    'x': x,
    'y': y
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
wasserstein_l1 Ran
ran on a synthesized input
microsoft/InfoGainalpharank/src/wasserstein/utils.py
code served (permissive licence) · get_code("1896b8300f6e8288")
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 float, 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 numpy as np

# Analyze the function:
# - base_dist: used with .shape[0], np.copy(), .clip(), and indexing [i]
#   so it's a 2D numpy array where first dimension is N
# - new_dist: used with .shape[0], np.copy(), .clip(), and operations like
#   np.abs(base_dist[i] - new_dist).sum(axis=1), so it's a 2D array
#   The subtraction suggests both have compatible shapes for broadcasting
# - normalise: boolean flag

# Create small test distributions
# N samples, each with some feature dimension
N = 2
feature_dim = 4

# base_dist: shape (N, feature_dim) - represents N samples with feature_dim features
base_dist = np.random.rand(N, feature_dim).astype(np.float64)

# new_dist: shape (N, feature_dim) - same structure
new_dist = np.random.rand(N, feature_dim).astype(np.float64)

# normalise: boolean
normalise = False

ARGS = {
    "base_dist": base_dist,
    "new_dist": new_dist,
    "normalise": normalise,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, ot — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, 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
alpha_rank Not yet run
unverified (harvested, no recorded run)
microsoft/InfoGainalpharank/src/alpha_rank.py
code served (permissive licence) · get_code("618c078adf9d0d72")
How it ran

no invocation recorded.

h_prob_func Not yet run
unverified (harvested, no recorded run)
microsoft/InfoGainalpharank/src/sampling_schemes/distributions/indep_normal.py
code served (permissive licence) · get_code("a382174b9540c1a9")
How it ran

no invocation recorded.

h_sample_prob_func Not yet run
unverified (harvested, no recorded run)
microsoft/InfoGainalpharank/src/sampling_schemes/distributions/indep_normal.py
code served (permissive licence) · get_code("d2f7c0b4bcba8784")
How it ran

no invocation recorded.

run_sampling Not yet run
unverified (harvested, no recorded run)
microsoft/InfoGainalpharank/src/sampling.py
code served (permissive licence) · get_code("85ebfeb89808327c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 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 9 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

This paper investigates the evaluation of learned multiagent strategies in the incomplete information setting, which plays a critical role in ranking and training of agents. Traditionally, researchers have relied on Elo ratings for this purpose, with recent works also using methods based on Nash equilibria. Unfortunately, Elo is unable to handle intransitive agent interactions, and other techniques are restricted to zero-sum, two-player settings or are limited by the fact that the Nash equilibrium is intractable to compute. Recently, a ranking method called α-Rank, relying on a new graph-based game-theoretic solution concept, was shown to tractably apply to general games. However, evaluations based on Elo or α-Rank typically assume noise-free game outcomes, despite the data often being collected from noisy simulations, making this assumption unrealistic in practice. This paper investigates multiagent evaluation in the incomplete information regime, involving general-sum many-player games with noisy outcomes. We derive sample complexity guarantees required to confidently rank agents in this setting. We propose adaptive algorithms for accurate ranking, provide correctness and sample complexity guarantees, then introduce a means of connecting uncertainties in noisy match outcomes to uncertainties in rankings. We evaluate the performance of these approaches in several domains, including Bernoulli games, a soccer meta-game, and Kuhn poker.

For agents

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

get_harvested_code_for_paper("1909.09849")
get_code_for_paper("1909.09849")
have("1909.09849")

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

curl https://syntology.ai/api/ran/1909.09849.json

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

[![Syntology run record](https://syntology.ai/api/ran/1909.09849.svg)](https://syntology.ai/paper/1909.09849)

Connect an agent — have() is free.