SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.14419 · NeurIPS · 2021

On the Estimation Bias in Double Q-Learning

Chongjie Zhang, Zhizhou Ren, Beining Han, Guangxiang Zhu, Hao Hu, Jianglun Chen

arXiv · PDF · Open in the Atlas

Code that ran

4 of 6 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 0 have no contract verdict; 3 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/2109.14419.json · badge: /api/ran/2109.14419.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
stilwell-git/doubly-bounded-q-learning canonical (classifier label)
classifier basis: classifier_2025
4 ran · 2 no recorded run
FunctionStatusWhere it lives
c_complie Ran
ran; our contract draft was wrong, not the code
stilwell-git/doubly-bounded-q-learning/algorithm/adp.py
code served (permissive licence) · get_code("d7c33d7c18e285de")
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 str. 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 os
import tempfile

# Create a temporary C file for testing
temp_dir = tempfile.gettempdir()
c_file_path = os.path.join(temp_dir, 'test_compile.c')

# Write a minimal valid C shared library
with open(c_file_path, 'w') as f:
    f.write('''
int add(int a, int b) {
    return a + b;
}
''')

ARGS = {
    'c_path': c_file_path,
    'so_path': None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
c_int Ran
ran, honoured the contract we drafted · constructed an object; no result was computed
stilwell-git/doubly-bounded-q-learning/algorithm/adp.py
code served (permissive licence) · get_code("e01bb91665e46e9b")
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, c_int: 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 ctypes

ARGS = {
    "value": 42
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type c_int — 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 c_int output the recorded ndim 0 and dtype family int 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
c_ptr Ran
ran; our contract draft was wrong, not the code
stilwell-git/doubly-bounded-q-learning/algorithm/adp.py
code served (permissive licence) · get_code("cfae173b4b9706ed")
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 c_char_p. 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 ctypes
import numpy as np
import torch

# Create a small numpy array for testing
np_array = np.array([1, 2, 3, 4, 5], dtype=np.float32)

ARGS = {
    "np_array": np_array
}
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 c_char_p — a c_char_p 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
cpp_complie Ran
ran; our contract draft was wrong, not the code
stilwell-git/doubly-bounded-q-learning/algorithm/adp.py
code served (permissive licence) · get_code("8f807e0093aa8b87")
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 str. 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 os
import tempfile

# Create a temporary C++ file for testing
temp_dir = tempfile.gettempdir()
cpp_path = os.path.join(temp_dir, 'test_compile.cpp')

# Write a minimal valid C++ file
with open(cpp_path, 'w') as f:
    f.write('''
#include <stdio.h>
extern "C" {
    int add(int a, int b) {
        return a + b;
    }
}
''')

so_path = os.path.join(temp_dir, 'test_compile.so')

ARGS = {
    'cpp_path': cpp_path,
    'so_path': so_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
AbstractedDynamicProgramming Not yet run
unverified (harvested, no recorded run)
stilwell-git/doubly-bounded-q-learning/algorithm/adp.py
code served (permissive licence) · get_code("268775a3310bf7e2")
How it ran

no invocation recorded.

load_c_lib Not yet run
unverified (harvested, no recorded run)
stilwell-git/doubly-bounded-q-learning/algorithm/adp.py
code served (permissive licence) · get_code("c5ad985be0dd4a68")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 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 6 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

Double Q-learning is a classical method for reducing overestimation bias, which is caused by taking maximum estimated values in the Bellman operation. Its variants in the deep Q-learning paradigm have shown great promise in producing reliable value prediction and improving learning performance. However, as shown by prior work, double Q-learning is not fully unbiased and suffers from underestimation bias. In this paper, we show that such underestimation bias may lead to multiple non-optimal fixed points under an approximate Bellman operator. To address the concerns of converging to non-optimal stationary solutions, we propose a simple but effective approach as a partial fix for the underestimation bias in double Q-learning. This approach leverages an approximate dynamic programming to bound the target value. We extensively evaluate our proposed method in the Atari benchmark tasks and demonstrate its significant improvement over baseline algorithms.

For agents

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

get_harvested_code_for_paper("2109.14419")
get_code_for_paper("2109.14419")
have("2109.14419")

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

curl https://syntology.ai/api/ran/2109.14419.json

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

[![Syntology run record](https://syntology.ai/api/ran/2109.14419.svg)](https://syntology.ai/paper/2109.14419)

Connect an agent — have() is free.