Overview
Field: Machine Learning, Reinforcement Learning, Game AI Authors: Derin Gezgin, Jim O'Connor, Tanner Goodwin Published: 2026-08-11 arXiv: 2508.03798
Summary
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.
- a random policy,
- an expert system,
- an evolutionary baseline, and
- PPO and DQN agents trained from visual input.
- The expert system and evolutionary baseline each defeat the tutorial boss Asylum Demon (peak win rates of 63% and 43%, respectively).
- None of the five methods defeat the other four bosses in DSLE-5.
- PPO and DQN show no measurable learning within the wall-clock budget already requiring tens of hours per run (up to 0.33% win rate on the tutorial boss, 0% elsewhere).
- A broader study running the evolutionary baseline across all 22 encounters, with overpowered level-50 stats, wins only a few additional early-game bosses; the rest remain unbeaten.
- Failure cases range from deaths in under 10 seconds on cramped, multi-target encounters to nearly minute-long stalemates with minimal damage dealt.
- Results are reported via survival time and damage dealt, not win rate alone, to capture partial progress.
We recommend DSLE-5 as the starting suite for agents built on DSLE.
On DSLE-5 we evaluate:
Key Findings
Original Abstract
> 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.