---
title: "Discussion on \"Why GRPO, Dr. GRPO, and DAPO Are All the Same Algorithm in Disguise: The Group-Standard-Deviation Identity\" | Hashnode"
description: "Discussion on \"Why GRPO, Dr. GRPO, and DAPO Are All the Same Algorithm in Disguise: The Group-Standard-Deviation Identity\". GRPO, Dr. GRPO, and DAPO all operate on sigma, the group reward std dev. Exact formulas for group size, silent-group waste, and difficulty bias."
url: "https://hashnode.com/posts/why-grpo-dr-grpo-and-dapo-are-all-the-same-algorithm-in-disguise-the-group-standard-deviation-identity/6a703395070bb5315cd59c36"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Why GRPO, Dr. GRPO, and DAPO Are All the Same Algorithm in Disguise: The Group-Standard-Deviation Identity](https://miainflorence.hashnode.dev/why-grpo-dr-grpo-and-dapo-are-all-the-same-algorithm-in-disguise-the-group-standard-deviation-identity?utm_source=hashnode&utm_medium=feed)

GRPO, Dr. GRPO, and DAPO all operate on sigma, the group reward std dev. Exact formulas for group size, silent-group waste, and difficulty bias.

Aug 3 · 9 min read · miainflorence.hashnode.dev  
Read the article

-   [#machine-learning](https://hashnode.com/tag/machine-learning)
-   [#reinforcement-learning](https://hashnode.com/tag/reinforcement-learning)
-   [#llm](https://hashnode.com/tag/llm)
-   [#deep-learning](https://hashnode.com/tag/deep-learning)

## Written by

[Chaeyeon Mia Lee](https://hashnode.com/@miainflorence)

@miainflorence · Seoul, South Korea

  
[View profile](https://hashnode.com/@miainflorence)

## Responses

Markdown supported · ⌘↵ to post

No responses yet.