SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1906.07975 · 2019

Batch Active Learning Using Determinantal Point Processes

arXiv · PDF · Open in the Atlas

Code that ran

7 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 7 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/1906.07975.json · badge: /api/ran/1906.07975.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
Stanford-ILIAD/DPP-Batch-Active-Learning canonical (classifier label)
classifier basis: census
7 ran · 2 no recorded run
FunctionStatusWhere it lives
feature Ran
ran on a synthesized input
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/feature.py
code served (permissive licence) · get_code("7b31c8b445431412")
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 Feature. 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 torch

torch.manual_seed(0)

def sample_function(x):
    return x * 2 + 1

ARGS = {
    'f': sample_function
}
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 Feature — a Feature output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Feature 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
func_psi Ran
ran on a synthesized input
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/algos.py
code served (permissive licence) · get_code("0af2862e2f7720fc")
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 1, dtype family float, shape (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

# Analyze func_psi:
# - psi_set: used with .dot(w_samples.T), so needs to be a 2D array (n_features x n_samples)
# - w_samples: used with .shape[0] and .T, so needs to be a 2D array (M x n_features)
# The function computes probabilities and entropy-like terms

# Create small, distinct-sized arrays
np.random.seed(0)

# w_samples: shape (M, n_features) where M=2, n_features=4
w_samples = np.random.randn(2, 4).astype(np.float64)

# psi_set: shape (n_basis, n_features) where n_basis=3, n_features=4
# This will produce output of shape (n_basis, M) = (3, 2)
psi_set = np.random.randn(3, 4).astype(np.float64)

ARGS = {
    "psi_set": psi_set,
    "w_samples": w_samples,
}
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 1, dtype family float, shape (3,) — shape (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
generate_psi Ran
ran on a synthesized input
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/algos.py
code served (permissive licence) · get_code("4497f69d2c6da5f1")
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 (2, 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 numpy as np

class SimulationObject:
    def __init__(self, feed_size, num_of_features):
        self.feed_size = feed_size
        self.num_of_features = num_of_features
        self._last_input = None
    
    def feed(self, input_list):
        """Store the input for later feature extraction"""
        self._last_input = np.array(input_list)
    
    def get_features(self):
        """Return dummy features based on the last input"""
        # Generate features as a function of input
        if self._last_input is not None:
            # Simple transformation: repeat and scale the input
            features = np.tile(self._last_input, 
                              (self.num_of_features // self.feed_size) + 1)
            return features[:self.num_of_features]
        return np.zeros(self.num_of_features)

# Create simulation object with feed_size=4 and num_of_features=5
simulation_object = SimulationObject(feed_size=4, num_of_features=5)

# Create inputs_set: 2D array with 2 samples, each sample has 2*feed_size=8 values
# (4 for inputs1 and 4 for inputs2)
# Using shape (2, 8) for 2 samples of 8 values each
inputs_set = np.array([
    [1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0],
    [9.0, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0, 16.0]
])

ARGS = {
    "simulation_object": simulation_object,
    "inputs_set": inputs_set
}
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 (2, 5) — shape (2, 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
kMedoids Ran
ran on a synthesized input
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/kmedoids.py
code served (permissive licence) · get_code("418285ba39c6240b")
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. 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

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

# Create a small distance matrix D
# D should be a 2D array where D[i,j] represents distance between points i and j
# Make it symmetric with zeros on diagonal (valid distance matrix)
# m rows (data points), n columns (also data points for medoid selection)
m, n = 6, 6
D = np.array([
    [0.0, 1.2, 2.3, 3.1, 2.8, 1.5],
    [1.2, 0.0, 1.9, 2.5, 2.1, 1.8],
    [2.3, 1.9, 0.0, 1.4, 3.2, 2.7],
    [3.1, 2.5, 1.4, 0.0, 2.9, 3.3],
    [2.8, 2.1, 3.2, 2.9, 0.0, 1.1],
    [1.5, 1.8, 2.7, 3.3, 1.1, 0.0]
], dtype=np.float64)

# Ensure symmetry
D = (D + D.T) / 2
np.fill_diagonal(D, 0.0)

# Parameters for kMedoids
k = 2  # Number of medoids
tmax = 100  # Maximum number of iterations

ARGS = {
    'D': D,
    'k': k,
    'tmax': tmax
}
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 tuple — a tuple output has no array shape of its own; none is 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
sample_ids_mc Ran
ran on a synthesized input
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/dpp_sampler.py
code served (permissive licence) · get_code("ec75197c05a8280e")
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 int; 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 test data
# points: list of coordinate tuples/arrays
points = [
    np.array([0.0, 0.0]),
    np.array([1.0, 1.0]),
    np.array([2.0, 2.0]),
    np.array([3.0, 3.0]),
    np.array([4.0, 4.0]),
]

# scores: array of positive values, one per point
scores = np.array([1.0, 2.0, 1.5, 0.8, 1.2])

# k: number of points to sample (greedy + mcmc)
k = 2

# alpha: temperature parameter for MCMC
alpha = 2.0

# gamma: exponent for score transformation
gamma = 1.0

# cond_ids: list of indices that must be included (conditioned on)
cond_ids = []

# steps: number of MCMC steps
steps = 100

ARGS = {
    "points": points,
    "scores": scores,
    "k": k,
    "alpha": alpha,
    "gamma": gamma,
    "cond_ids": cond_ids,
    "steps": steps,
}
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 list, ndim 1, dtype family int — 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
sample_mc Ran
ran on a synthesized input
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/dpp_sampler.py
code served (permissive licence) · get_code("ba6d8c4f3f2eae3d")
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 (2, 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
from scipy.spatial.distance import cdist

# Create small synthetic data
np.random.seed(0)

# Define points in a small 3D space
points = [
    np.array([0.0, 0.0, 0.0]),
    np.array([1.0, 0.0, 0.0]),
    np.array([0.0, 1.0, 0.0]),
    np.array([1.0, 1.0, 0.0]),
    np.array([0.5, 0.5, 1.0]),
]

# Define scores for each point
scores = np.array([1.0, 0.8, 0.9, 0.7, 0.95])

# Parameters
k = 2  # number of points to sample
alpha = 2.0
gamma = 1.0
cond_ids = []  # no conditioning ids

ARGS = {
    "points": points,
    "scores": scores,
    "k": k,
    "alpha": alpha,
    "gamma": gamma,
    "cond_ids": cond_ids,
}
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 ndarray, ndim 2, dtype family float, shape (2, 3) — shape (2, 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
speed Ran
ran on a synthesized input
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/feature.py
code served (permissive licence) · get_code("bbfd88d6d34e2002")
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 Feature. 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 torch
import numpy as np

torch.manual_seed(0)

class Feature(object):

    def __init__(self, f):
        self.f = f

    def __call__(self, *args):
        return self.f(*args)

    def __add__(self, r):
        return Feature(lambda *args: self(*args) + r(*args))

    def __radd__(self, r):
        return Feature(lambda *args: r(*args) + self(*args))

    def __mul__(self, r):
        return Feature(lambda *args: self(*args) * r)

    def __rmul__(self, r):
        return Feature(lambda *args: r * self(*args))

    def __pos__(self, r):
        return self

    def __neg__(self):
        return Feature(lambda *args: -self(*args))

    def __sub__(self, r):
        return Feature(lambda *args: self(*args) - r(*args))

    def __rsub__(self, r):
        return Feature(lambda *args: r(*args) - self(*args))

def feature(f):
    return Feature(f)

def speed(s=1.0):

    @feature
    def f(t, x, u):
        return -(x[3] - s) * (x[3] - s)
    return f

# Setup for testing
# The speed function takes one parameter: s (default 1.0)
# The returned feature function expects: t, x, u
# where x is indexed with [3], so x must be indexable with at least 4 elements

t_val = 0.5
x_val = torch.tensor([1.0, 2.0, 3.0, 4.5])
u_val = torch.tensor([0.1, 0.2])

ARGS = {
    "s": 1.0
}
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 Feature — a Feature output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Feature 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
func Not yet run
unverified (harvested, no recorded run)
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/algos.py
code served (permissive licence) · get_code("36a4e0da182e8207")
How it ran

no invocation recorded.

setup_sampler Not yet run
unverified (harvested, no recorded run)
Stanford-ILIAD/DPP-Batch-Active-Learning/reward_learning/dpp_sampler.py
code served (permissive licence) · get_code("bfc97faec4a3b7cb")
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, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 2 later) and 4 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
feature 3 other papers
1 dated earlier, 2 later
1810.04303 (2018-10)
Batch Active Preference-Based Learning of Reward Functions
1910.04365 (2019-10)
Asking Easy Questions: A User-Friendly Approach to Active Reward Learning
2110.00284 (2021-10)
Learning Reward Functions from Scale Feedback
speed 3 other papers
1 dated earlier, 2 later
1810.04303 (2018-10)
Batch Active Preference-Based Learning of Reward Functions
1910.04365 (2019-10)
Asking Easy Questions: A User-Friendly Approach to Active Reward Learning
2110.00284 (2021-10)
Learning Reward Functions from Scale Feedback
func 1 other paper
1 dated earlier
1810.04303 (2018-10)
Batch Active Preference-Based Learning of Reward Functions
generate_psi 1 other paper
1 dated earlier
1810.04303 (2018-10)
Batch Active Preference-Based Learning of Reward Functions
kMedoids 1 other paper
1 dated earlier
1810.04303 (2018-10)
Batch Active Preference-Based Learning of Reward Functions

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

Data collection and labeling is one of the main challenges in employing machine learning algorithms in a variety of real-world applications with limited data. While active learning methods attempt to tackle this issue by labeling only the data samples that give high information, they generally suffer from large computational costs and are impractical in settings where data can be collected in parallel. Batch active learning methods attempt to overcome this computational burden by querying batches of samples at a time. To avoid redundancy between samples, previous works rely on some ad hoc combination of sample quality and diversity. In this paper, we present a new principled batch active learning method using Determinantal Point Processes, a repulsive point process that enables generating diverse batches of samples. We develop tractable algorithms to approximate the mode of a DPP distribution, and provide theoretical guarantees on the degree of approximation. We further demonstrate that an iterative greedy method for DPP maximization, which has lower computational costs but worse theoretical guarantees, still gives competitive results for batch active learning. Our experiments show the value of our methods on several datasets against state-of-the-art baselines.

For agents

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

get_harvested_code_for_paper("1906.07975")
get_code_for_paper("1906.07975")
have("1906.07975")

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

curl https://syntology.ai/api/ran/1906.07975.json

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

[![Syntology run record](https://syntology.ai/api/ran/1906.07975.svg)](https://syntology.ai/paper/1906.07975)

Connect an agent — have() is free.