SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2301.00557 · ICML · 2023

Learning to Maximize Mutual Information for Dynamic Feature Selection

Ian Covert, Su-In Lee, Wei Qiu, Mingyu Lu, Nayoon Kim, Nathan White

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 failures of our instrument; 1 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/2301.00557.json · badge: /api/ran/2301.00557.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
iancovert/dynamic-selection canonical (classifier label)
classifier basis: classifier_2025
3 ran · 1 no recorded run
FunctionStatusWhere it lives
ConcreteSelector Ran
ran on a synthesized input
iancovert/dynamic-selection/dynamic_selection/greedy.py
code served (permissive licence) · get_code("233c9ebbe363a9cc")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, 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 torch
import torch.nn as nn

INIT_ARGS = {
    'gamma': 0.2
}

ARGS = {
    'logits': torch.randn(2, 5),
    'temp': torch.tensor(1.0),
    'deterministic': False
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
GreedyDynamicSelection Ran
ran on a synthesized input · constructed an object; no result was computed
iancovert/dynamic-selection/dynamic_selection/greedy.py
code served (permissive licence) · get_code("158e57a1e13ee09d")
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 an instance of its own class, GreedyDynamicSelection: the run constructed the object and did not call it, so no result was computed. 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 torch.nn as nn
import numpy as np

torch.manual_seed(0)

# Define a simple selector model
class SimpleSelector(nn.Module):
    def __init__(self, input_size, output_size):
        super().__init__()
        self.fc = nn.Linear(input_size, output_size)
    
    def forward(self, x):
        return self.fc(x)

# Define a simple predictor model
class SimplePredictor(nn.Module):
    def __init__(self, input_size, output_size):
        super().__init__()
        self.fc = nn.Linear(input_size, output_size)
    
    def forward(self, x):
        return self.fc(x)

# Define a simple mask layer
class SimpleMaskLayer(nn.Module):
    def __init__(self, mask_size=None):
        super().__init__()
        self.mask_size = mask_size
    
    def forward(self, x, m):
        # Element-wise multiplication to apply mask
        return x * m

# Create small sample data loaders
def create_data_loader(batch_size=2, num_features=4, num_samples=4, output_size=2):
    data = []
    for _ in range(num_samples):
        x = torch.randn(batch_size, num_features)
        y = torch.randint(0, output_size, (batch_size,))
        data.append((x, y))
    return data

train_loader = create_data_loader(batch_size=2, num_features=4, num_samples=3, output_size=2)
val_loader = create_data_loader(batch_size=2, num_features=4, num_samples=2, output_size=2)

# Instantiate models
num_features = 4
output_size = 2
selector = SimpleSelector(input_size=num_features, output_size=num_features)
predictor = SimplePredictor(input_size=num_features, output_size=output_size)
mask_layer = SimpleMaskLayer(mask_size=num_features)

# Simple loss function
loss_fn = nn.CrossEntropyLoss()

ARGS = {
    'selector': selector,
    'predictor': predictor,
    'mask_layer': mask_layer,
}
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 GreedyDynamicSelection — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a GreedyDynamicSelection 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
make_onehot Ran
ran, honoured the contract we drafted
iancovert/dynamic-selection/dynamic_selection/greedy.py
code served (permissive licence) · get_code("f8d153b8569b0e13")
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)

# Create a small approximately one-hot tensor
# Shape: (batch_size=2, num_classes=4)
# The tensor should have approximately one-hot structure where one element
# per row is much larger than others
x = torch.tensor([
    [0.1, 0.8, 0.05, 0.05],
    [0.2, 0.1, 0.6, 0.1]
], dtype=torch.float32)

ARGS = {
    "x": x
}
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 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
restore_parameters Not yet run
unverified (harvested, no recorded run)
iancovert/dynamic-selection/dynamic_selection/greedy.py
code served (permissive licence) · get_code("01e0b17ac1cac60a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 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.

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

Feature selection helps reduce data acquisition costs in ML, but the standard approach is to train models with static feature subsets. Here, we consider the dynamic feature selection (DFS) problem where a model sequentially queries features based on the presently available information. DFS is often addressed with reinforcement learning, but we explore a simpler approach of greedily selecting features based on their conditional mutual information. This method is theoretically appealing but requires oracle access to the data distribution, so we develop a learning approach based on amortized optimization. The proposed method is shown to recover the greedy policy when trained to optimality, and it outperforms numerous existing feature selection methods in our experiments, thus validating it as a simple but powerful approach for this problem.

For agents

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

get_harvested_code_for_paper("2301.00557")
get_code_for_paper("2301.00557")
have("2301.00557")

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

curl https://syntology.ai/api/ran/2301.00557.json

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

[![Syntology run record](https://syntology.ai/api/ran/2301.00557.svg)](https://syntology.ai/paper/2301.00557)

Connect an agent — have() is free.