GPT-2 Architecture and Training Details: Parameters & Cross-Entropy Loss

cover

24 Jun 2025

Table of Links

Abstract and 1 Introduction

3 Model and 3.1 Associative memories

3.2 Transformer blocks

4 A New Energy Function

4.1 The layered structure

5 Cross-Entropy Loss

6 Empirical Results and 6.1 Empirical evaluation of the radius

6.2 Training GPT-2

6.3 Training Vanilla Transformers

7 Conclusion and Acknowledgments

Appendix A. Deferred Tables

Appendix B. Some Properties of the Energy Functions

Appendix C. Deferred Proofs from Section 5

Appendix D. Transformer Details: Using GPT-2 as an Example

Appendix D. Transformer Details: Using GPT-2 as an Example

Another commonly used loss is the perplexity, which is equivalent to the exponentiated version of the cross-entropy.

Authors:

(1) Xueyan Niu, Theory Laboratory, Central Research Institute, 2012 Laboratories, Huawei Technologies Co., Ltd.;

(2) Bo Bai baibo (8@huawei.com);

(3) Lei Deng (deng.lei2@huawei.com);

(4) Wei Han (harvey.hanwei@huawei.com).

This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license.

Theoretical Derivations: Cross-Entropy Loss and Energy Functions in LLMs

Transformer Theory & LLM References: Here's What You Should Check Out