DSLE: A Learning Environment for Dark Souls Boss Encounters

2026-08-10

AI Summary

This work addresses the challenge of learning effective combat strategies in realistic action games with high-dimensional visual inputs and sparse rewards. The authors propose DSLE, the first containerized platform that systematically converts all 22 boss fights from *Dark Souls: Remastered* into Gymnasium-compatible environments, and introduce DSLE-5—a benchmark suite comprising five representative scenarios. The platform supports training via pixel inputs using mainstream methods such as PPO, DQN, evolutionary algorithms, and expert systems. Experimental results show that only expert systems and evolutionary algorithms achieve non-trivial success against the tutorial boss (peak win rate: 63%), while standard deep reinforcement learning approaches fail to learn effectively. Even with a max-level character, existing methods struggle to defeat most bosses, highlighting the substantial difficulty of policy learning in authentic game environments.

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. 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.

Problem

• Dark Souls
• reinforcement learning
• sparse rewards
• game AI
• boss encounters

Innovation

• Dark Souls Learning Environment
• sparse rewards
• high-dimensional visual input
• real-time game interaction
• reinforcement learning benchmark