---
title: Game AI Evolution
url: https://www.ml-quant.com/papers/ssrn/4772907/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: SSRN 4772907
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4772907
featured: 2024-03-27
citations: unknown
topic: ML & AI Methods
---


# Game AI Evolution

The research investigates the use of deep reinforcement learning in creating adaptable and compliant gaming AI, overcoming limitations of traditional methods.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4772907
- Identifier: SSRN 4772907
- Released: 2024-03-26
- First featured: Quant Letter No. 42 (2024-03-27): https://www.ml-quant.com/issues/2024-03-27/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: ML & AI Methods

## Related

- [Mastering Diverse Domains through World Models](https://www.ml-quant.com/papers/arxiv/2301.04104/): Algorithm Mastery: DreamerV3, a universal algorithm, excels in over 150 varied tasks, including diamond collection in Minecraft without human input, expanding the scope of reinforcement learning.
- [SimPO: Simple Preference Optimization with a Reference-Free Reward](https://www.ml-quant.com/papers/arxiv/2405.14734/): Simple Preference Optimization: SimPO improves reinforcement learning from human feedback by using the average log probability of a sequence as the implicit reward, enhancing training stability and computational efficiency.
- [Settling the Sample Complexity of Model-Based Offline Reinforcement Learning](https://www.ml-quant.com/papers/arxiv/2204.05275/): A paper reveals that a model-based approach can achieve optimal sample complexity without burn-in cost in offline reinforcement learning for tabular Markov decision processes, providing an efficient solution for sample-starved applications.
- [DPO Meets PPO: Reinforced Token Optimization for RLHF](https://www.ml-quant.com/papers/arxiv/2404.18922/): A new framework is introduced that models Reinforcement Learning from Human Feedback as a Markov decision process, using an algorithm that learns from preference data.
- [Critique-out-Loud Reward Models](https://www.ml-quant.com/papers/arxiv/2408.11791/): The article presents CLoud reward models that use human feedback to improve reinforcement learning, enhancing accuracy and win rate in ArenaHard.
- [Transformers as Decision Makers: Provable In-Context Reinforcement Learning via Supervised Pretraining](https://www.ml-quant.com/papers/arxiv/2310.08566/): The article presents a theoretical framework for training large transformer models for in-context reinforcement learning, offering the first quantitative analysis of their capabilities.