# Odyssey — Introducing PROWL-2: Jointly Learning Simulation and Decision-Making

- Company: Odyssey (odyssey.systems)
- Announced: 2026-10-01
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://odyssey.systems/introducing-prowl-2
- Record: https://forck.live/items/18588-introducing-prowl-2-jointly-learning-simulation-and-decision-making
- Subject: Odyssey world models

Odyssey introduced PROWL-2, a framework for multi-agent learning from imagined experience that jointly trains a team of agents and its world model using separate curricula within a single training loop. It combines task-policy curriculum learning, world-model curriculum learning, and a fidelity gate to filter hallucinated rollouts. In evaluations on SMACv2 and MQE, PROWL-2 achieved the highest mean performance in all nine SMACv2 scenarios and improved success on the hardest MQE tasks from 29.8% to 70.4% on Gate-3 and from 7.3% to 28.2% on Shepherd-Hard.

## Evidence

Verbatim from https://odyssey.systems/introducing-prowl-2:

> PROWL-2 is the first framework in which a team of agents and its world model each learn from their own curriculum, continually and within a single training loop.

---

Record: https://forck.live/items/18588-introducing-prowl-2-jointly-learning-simulation-and-decision-making
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md