mlsanityData to 14 Oct 2026

Terra Nova: A Comprehensive Challenge Environment for Intelligent Agents

arXiv 2511.15378 ↗First posted 19 Nov 2025cs.AIcited by 0

On the map: Deep reinforcement learning methods → RL training of language and agent models

Abstract (arXiv)

We introduce Terra Nova, a new comprehensive challenge environment (CCE) for reinforcement learning (RL) research inspired by Civilization V. A CCE is a single environment in which multiple canonical RL challenges (e.g., partial observability, credit assignment, representation learning, enormous action spaces, etc.) arise simultaneously. Mastery therefore demands integrated, long-horizon understanding across many interacting variables. We emphasize that this definition excludes challenges that only aggregate unrelated tasks in independent, parallel streams (e.g., learning to play all Atari games at once). These aggregated multitask benchmarks primarily asses whether an agent can catalog and switch among unrelated policies rather than test an agent's ability to perform deep reasoning across many interacting challenges.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL training of language and agent models, one of 15 directions in Deep reinforcement learning methods. Builds on 8 of its 12 references that are in the map. First posted on arXiv 19 Nov 2025.

Builds on

8 of 12 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement Learning26 Feb

Other references · 7

  1. 2019Dota 2 with Large Scale Deep Reinforcement Learningcited by 1,103
  2. 2020The NetHack Learning Environmentcited by 198
  3. 2019OpenSpiel: A Framework for Reinforcement Learning in Gamescited by 186
  4. 2019Neural MMO: A Massively Multiagent Game Environment for Training and Evaluating Intelligent Agentscited by 44
  5. 2024CivRealm: A Learning and Reasoning Odyssey in Civilization for Decision-Making Agentscited by 42
  6. 2019No Press Diplomacy: Modeling Multi-Agent Gameplaycited by 38
  7. 2024Towards General Negotiation Strategies with End-to-End Reinforcement Learningcited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Learning to Explore: Scaling Agentic Reasoning via Exploration-Aware Policy Optimization
  2. 2026Beyond Reward Engineering: A Data Recipe for Long-Context Reinforcement Learning
  3. 2026Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence
  4. 2026Skill0.5: Joint Skill Internalization and Utilization for Out-of-Distribution Generalization in Agentic Reinforcement Learning
  5. 2026MARS2: Scaling Multi-Agent Tree Search via Reinforcement Learning for Code Generation
  6. 2026LUMINA: Long-horizon Understanding for Multi-turn Interactive Agents
  7. 2026SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization
  8. 2026SkillRise: Agentic Reinforcement Learning for Cross-Task Skill Evolution