On the Mistaken Assumption of Interchangeable Deep Reinforcement Learning Implementations

• URL: http://arxiv.org/abs/2503.22575v1
• Date: Fri, 28 Mar 2025 16:25:06 GMT
• Title: On the Mistaken Assumption of Interchangeable Deep Reinforcement Learning Implementations
• Authors: Rajdeep Singh Hundal, Yan Xiao, Xiaochun Cao, Jin Song Dong, Manuel Rigger,
• Abstract summary: Deep Reinforcement Learning (DRL) is a paradigm of artificial intelligence where an agent uses a neural network to learn which actions to take in a given environment.<n>DRL has recently gained traction from being able to solve complex environments like driving simulators, 3D robotic control, and multiplayer-online-battle-arena video games.<n>Numerous implementations of the state-of-the-art algorithms responsible for training these agents, like the Deep Q-Network (DQN) and Proximal Policy Optimization (PPO) algorithms, currently exist.
• Score: 53.0667196725616
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Deep Reinforcement Learning (DRL) is a paradigm of artificial intelligence where an agent uses a neural network to learn which actions to take in a given environment. DRL has recently gained traction from being able to solve complex environments like driving simulators, 3D robotic control, and multiplayer-online-battle-arena video games. Numerous implementations of the state-of-the-art algorithms responsible for training these agents, like the Deep Q-Network (DQN) and Proximal Policy Optimization (PPO) algorithms, currently exist. However, studies make the mistake of assuming implementations of the same algorithm to be consistent and thus, interchangeable. In this paper, through a differential testing lens, we present the results of studying the extent of implementation inconsistencies, their effect on the implementations' performance, as well as their impact on the conclusions of prior studies under the assumption of interchangeable implementations. The outcomes of our differential tests showed significant discrepancies between the tested algorithm implementations, indicating that they are not interchangeable. In particular, out of the five PPO implementations tested on 56 games, three implementations achieved superhuman performance for 50% of their total trials while the other two implementations only achieved superhuman performance for less than 15% of their total trials. As part of a meticulous manual analysis of the implementations' source code, we analyzed implementation discrepancies and determined that code-level inconsistencies primarily caused these discrepancies. Lastly, we replicated a study and showed that this assumption of implementation interchangeability was sufficient to flip experiment outcomes. Therefore, this calls for a shift in how implementations are being used.

Related papers

• Exploring and Lifting the Robustness of LLM-powered Automated Program Repair with Metamorphic Testing [31.327835928133535]
  Large language model-powered Automated Program Repair (LAPR) techniques have achieved state-of-the-art bug-fixing performance.<n>It is crucial to conduct robustness testing on LAPR techniques before their practical deployment.<n>We propose MT-LAPR, a Metamorphic Testing framework exclusively for LAPR techniques.
  arXiv  Detail & Related papers  (2024-10-10T01:14:58Z)
• Bidirectional Decoding: Improving Action Chunking via Closed-Loop Resampling [51.38330727868982]
  Bidirectional Decoding (BID) is a test-time inference algorithm that bridges action chunking with closed-loop operations.<n>We show that BID boosts the performance of two state-of-the-art generative policies across seven simulation benchmarks and two real-world tasks.
  arXiv  Detail & Related papers  (2024-08-30T15:39:34Z)
• Is Value Functions Estimation with Classification Plug-and-play for Offline Reinforcement Learning? [1.9116784879310031]
  In deep Reinforcement Learning (RL), value functions are approximated using deep neural networks and trained via mean squared error regression objectives. Recent research has proposed an alternative approach, utilizing the cross-entropy classification objective. Our work seeks to empirically investigate the impact of such a replacement in an offline RL setup.
  arXiv  Detail & Related papers  (2024-06-10T14:25:11Z)
• Planning for Sample Efficient Imitation Learning [52.44953015011569]
  Current imitation algorithms struggle to achieve high performance and high in-environment sample efficiency simultaneously. We propose EfficientImitate, a planning-based imitation learning method that can achieve high in-environment sample efficiency and performance simultaneously. Experimental results show that EI achieves state-of-the-art results in performance and sample efficiency.
  arXiv  Detail & Related papers  (2022-10-18T05:19:26Z)
• Actor Prioritized Experience Replay [0.0]
  Prioritized Experience Replay (PER) allows agents to learn from transitions sampled with non-uniform probability proportional to their temporal-difference (TD) error. We introduce a novel experience replay sampling framework for actor-critic methods, which also regards issues with stability and recent findings behind the poor empirical performance of PER. An extensive set of experiments verifies our theoretical claims and demonstrates that the introduced method significantly outperforms the competing approaches.
  arXiv  Detail & Related papers  (2022-09-01T15:27:46Z)
• Benchopt: Reproducible, efficient and collaborative optimization benchmarks [67.29240500171532]
  Benchopt is a framework to automate, reproduce and publish optimization benchmarks in machine learning. Benchopt simplifies benchmarking for the community by providing an off-the-shelf tool for running, sharing and extending experiments.
  arXiv  Detail & Related papers  (2022-06-27T16:19:24Z)
• Benchmarking Deep Models for Salient Object Detection [67.07247772280212]
  We construct a general SALient Object Detection (SALOD) benchmark to conduct a comprehensive comparison among several representative SOD methods. In the above experiments, we find that existing loss functions usually specialized in some metrics but reported inferior results on the others. We propose a novel Edge-Aware (EA) loss that promotes deep networks to learn more discriminative features by integrating both pixel- and image-level supervision signals.
  arXiv  Detail & Related papers  (2022-02-07T03:43:16Z)
• Can Active Learning Preemptively Mitigate Fairness Issues? [66.84854430781097]
  dataset bias is one of the prevailing causes of unfairness in machine learning. We study whether models trained with uncertainty-based ALs are fairer in their decisions with respect to a protected class. We also explore the interaction of algorithmic fairness methods such as gradient reversal (GRAD) and BALD.
  arXiv  Detail & Related papers  (2021-04-14T14:20:22Z)
• DERAIL: Diagnostic Environments for Reward And Imitation Learning [9.099589602551573]
  We develop a suite of diagnostic tasks that test individual facets of algorithm performance in isolation. Results confirm that algorithm performance is highly sensitive to implementation details. Case-study shows how the suite can pinpoint design flaws and rapidly evaluate candidate solutions.
  arXiv  Detail & Related papers  (2020-12-02T18:07:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.