SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2005.12729 · ICLR · 2020

Implementation Matters in Deep Policy Gradients: A Case Study on PPO and TRPO

Logan Engstrom, Andrew Ilyas, Shibani Santurkar, Dimitris Tsipras, Firdaus Janoos, Larry Rudolph, Aleksander Ądry, Two Sigma

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 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, 0 violated it and 0 have no contract verdict; 1 failure of our instrument; 0 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/2005.12729.json · badge: /api/ran/2005.12729.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
MadryLab/implementation-matters canonical (classifier label)
classifier basis not recorded on the link
2 ran · 0 no recorded run
FunctionStatusWhere it lives
dict_product Ran
ran; our contract draft was wrong, not the code
MadryLab/implementation-matters/src/utils.py
code served (permissive licence) · get_code("35b6f17122388f57")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned map. 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 itertools

# The function expects a dictionary where each value is a list of possible values
# It will create all combinations using itertools.product

d = {
    "learning_rate": [0.001, 0.01, 0.1],
    "batch_size": [16, 32],
    "optimizer": ["adam", "sgd"],
    "num_layers": [2, 3, 4]
}

ARGS = {"d": d}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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 map — a map output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a map 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
iwt Ran
ran, honoured the contract we drafted
MadryLab/implementation-matters/src/utils.py
code served (permissive licence) · get_code("942706632c22fc44")
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

ARGS = {
    "start": 0,
    "end": 10,
    "interval": 2,
    "trials": 3
}
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 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

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 0 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
dict_product 4 other papers
4 later
2009.10897 (2020-09)
Revisiting Design Choices in Proximal Policy Optimization
2302.06912 (2023-02)
Regret-Based Defense in Adversarial Reinforcement Learning
2406.04724 (2024-06)
ON MINIMIZING ADVERSARIAL COUNTERFACTUAL ER-ROR IN ADVERSARIAL RL
and 1 more not named here
iwt 3 other papers
3 later
2009.10897 (2020-09)
Revisiting Design Choices in Proximal Policy Optimization
2302.06912 (2023-02)
Regret-Based Defense in Adversarial Reinforcement Learning
2406.04724 (2024-06)
ON MINIMIZING ADVERSARIAL COUNTERFACTUAL ER-ROR IN ADVERSARIAL RL

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 study the roots of algorithmic progress in deep policy gradient algorithms through a case study on two popular algorithms: Proximal Policy Optimization (PPO) and Trust Region Policy Optimization (TRPO). Specifically, we investigate the consequences of "code-level optimizations:" algorithm augmentations found only in implementations or described as auxiliary details to the core algorithm. Seemingly of secondary importance, such optimizations turn out to have a major impact on agent behavior. Our results show that they (a) are responsible for most of PPO's gain in cumulative reward over TRPO, and (b) fundamentally change how RL methods function. These insights show the difficulty and importance of attributing performance gains in deep reinforcement learning. * Equal contribution. Work done in part while interning at Two Sigma.

For agents

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

get_harvested_code_for_paper("2005.12729")
get_code_for_paper("2005.12729")
have("2005.12729")

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

curl https://syntology.ai/api/ran/2005.12729.json

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

[![Syntology run record](https://syntology.ai/api/ran/2005.12729.svg)](https://syntology.ai/paper/2005.12729)

Connect an agent — have() is free.