Gradient Estimation Using Stochastic Computation Graphs
Authors: John Schulman, Nicolas Heess, Theophane Weber, Pieter Abbeel
NeurIPS 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We introduce the formalism of stochastic computation graphs directed acyclic graphs that include both deterministic functions and conditional probability distributions and describe how to easily and automatically derive an unbiased estimator of the loss function s gradient. The resulting algorithm for computing the gradient estimator is a simple modiļ¬cation of the standard backpropagation algorithm. The contributions of this work are as follows: We introduce a formalism of stochastic computation graphs, and in this general setting, we derive unbiased estimators for the gradient of the expected loss. |
| Researcher Affiliation | Collaboration | 1 Google Deep Mind 2 University of California, Berkeley, EECS Department |
| Pseudocode | Yes | Algorithm 1 Compute Gradient Estimator for Stochastic Computation Graph |
| Open Source Code | No | Insufficient information. The paper describes a theoretical framework and algorithm but does not provide concrete access to source code for its implementation. |
| Open Datasets | No | Insufficient information. The paper is theoretical and does not report on experiments, so it does not mention the use of any publicly available datasets for training. |
| Dataset Splits | No | Insufficient information. The paper is theoretical and does not report on experiments, therefore it does not provide specific dataset split information for training, validation, or testing. |
| Hardware Specification | No | Insufficient information. The paper describes a theoretical framework and does not report on experiments, thus no specific hardware details are mentioned. |
| Software Dependencies | No | Insufficient information. The paper describes a theoretical framework and does not report on experiments or provide an implementation, thus no specific ancillary software details with version numbers are mentioned. |
| Experiment Setup | No | Insufficient information. The paper is theoretical and does not report on experiments, thus it does not provide specific experimental setup details such as hyperparameters or training configurations. |