Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note4 min read

DSLE: A Learning Environment for Dark Souls Boss Encounters

Source

arXiv

arxiv.org/abs/2608.09902v1

Read the source

Source

  • Source name: arXiv
  • Source kind: paper
  • Authors / organization: Derin Gezgin, Jim O'Connor, Tanner Goodwin, Gary B. Parker
  • Venue: arXiv: cs.AI
  • Published: 2026-08-10
  • Collected: 2026-08-11
  • URL: http://arxiv.org/abs/2608.09902v1
  • PDF: https://arxiv.org/pdf/2608.09902v1
  • DOI: UNKNOWN
  • Registry group: academic_apis_and_preprints
  • Evidence role: canonical_academic_index
  • Language: en
  • Country / region: UNKNOWN

Plain-Language Brief

Researchers have created a learning environment for AI agents called DSLE, which is based on the game Dark Souls: Remastered. The environment includes 22 boss encounters that can be used as benchmarks. The researchers define a subset of five bosses (DSLE-5) to start with and evaluate various AI methods like a random policy, an expert system, an evolutionary baseline, PPO, and DQN agents on this subset. They find that while the expert system and evolutionary baseline can beat the tutorial boss, none of the methods can defeat the other four bosses in DSLE-5. The study also shows that the AI struggles to learn on the DSLE-5 environment, taking tens of hours per run with little progress.

What the Document Says

  • DSLE is a learning environment for Dark Souls: Remastered
  • DSLE contains 22 boss encounters
  • DSLE is presented as game-playing agent benchmarks through Gymnasium-style interface
  • DSLE includes real-time combat, high-dimensional visual input, and sparse terminal rewards
  • Each environment step in DSLE is a real action executed against the running game
  • DSLE-5 is defined, containing melee fight, spatially constrained arena, environmental-hazard fight, multi-target fight, and fast final-boss fight
  • Expert system and evolutionary baseline can defeat Asylum Demon (the tutorial boss) but not other DSLE-5 bosses
  • PPO and DQN show no measurable learning within a budget that costs tens of wall-clock hours per run on DSLE-5
  • A broader study running the evolutionary baseline across all 22 encounters under advantaged all level-50 stats yields wins on only a handful of additional early-game bosses and leaves the rest unwon

Key Concepts Mentioned

  • AI agents
  • learning environment
  • Dark Souls: Remastered
  • DSLE-5
  • PPO
  • DQN

Claims Made by the Document

  • DSLE can be used as a benchmark for game-playing AI agents
  • The expert system and evolutionary baseline can defeat the Asylum Demon
  • None of the methods (expert system, evolutionary baseline, PPO, DQN) can defeat the other four DSLE-5 bosses
  • PPO and DQN struggle to learn on DSLE-5
  • A broader study shows that the AI struggles to beat early-game bosses outside of DSLE-5

Evidence Used

  • The text mentions DSLE being a learning environment for Dark Souls: Remastered and containing 22 boss encounters
  • DSLE-5 is defined and evaluated with various AI methods in the text
  • Win rates of the expert system, evolutionary baseline, PPO, and DQN on DSLE-5 are mentioned
  • The struggle of PPO and DQN to learn on DSLE-5 is implied by their low win rates

Methods / Approach

  • Expert system
  • evolutionary baseline
  • PPO
  • DQN

Limitations Stated by the Authors

The study only evaluates a limited number of AI methods and does not test them extensively across all 22 boss encounters in DSLE

Limitations Noted by the Agent

Agent interpretation — not from the source.

The agent may not be able to fully understand the game mechanics or strategies required to defeat certain bosses in DSLE, resulting in difficulty learning and winning

Open Questions

  • What are the specific challenges that AI agents face when trying to learn and beat the bosses in DSLE?
  • Can other AI methods perform better than those evaluated in this study on DSLE?

Notes for Future Q&A

May include interpretation. Labeled where it does.

The text does not provide detailed explanations for why certain AI methods struggle to learn and win on DSLE. It also does not compare the performance of the methods against each other or against a human player.

Raw Abstract / Excerpt

We introduce the Dark Souls Learning Environment (DSLE), a containerized platform that presents all 22 boss encounters of Dark Souls: Remastered as game-playing agent benchmarks through a Gymnasium-style interface. DSLE combines real-time combat, high-dimensional visual input, and sparse terminal rewards, with each environment step being a real action executed against the running game. To support controlled comparison, we define DSLE-5, a representative five-boss subset, spanning a melee fight, a spatially constrained arena, an environmental-hazard fight, a multi-target fight, and a fast final-boss fight, that we recommend as the starting suite for agents built on DSLE. On DSLE-5 we evaluate a random policy, an expert system, an evolutionary baseline, and PPO and DQN agents trained from visual input. The expert system and the evolutionary baseline each defeat the Asylum Demon, the game's tutorial boss (63% and 43% peak win rates), but none of the five methods defeats the other four DSLE-5 bosses; PPO and DQN show no measurable learning (at most 0.33% win rate on the tutorial boss, 0% elsewhere) within a budget that already costs tens of wall-clock hours per run. A broader study running the evolutionary baseline across all 22 encounters under advantaged all level-50 stats yields wins on only a handful of additional early-game bosses and leaves the rest unwon. The failure cases range from sub-10-second deaths in cramped, multi-target encounters to minute-long stalemates that inflict almost no damage, and we report them through survival time and damage dealt rather than win rate alone.