Self-Guided Plan Extraction for Instruction-Following Tasks with Goal-Conditional Reinforcement Learning
This addresses the problem of reducing manual effort in instruction-following for AI agents, though it appears incremental as it builds on prior methods with iterative co-training.
The paper tackles instruction-following tasks by introducing SuperIgor, a framework that uses a self-learning mechanism to generate and refine high-level plans, reducing manual annotation needs. Results show that SuperIgor agents adhere to instructions more strictly than baselines and generalize well to unseen instructions.
We introduce SuperIgor, a framework for instruction-following tasks. Unlike prior methods that rely on predefined subtasks, SuperIgor enables a language model to generate and refine high-level plans through a self-learning mechanism, reducing the need for manual dataset annotation. Our approach involves iterative co-training: an RL agent is trained to follow the generated plans, while the language model adapts and modifies these plans based on RL feedback and preferences. This creates a feedback loop where both the agent and the planner improve jointly. We validate our framework in environments with rich dynamics and stochasticity. Results show that SuperIgor agents adhere to instructions more strictly than baseline methods, while also demonstrating strong generalization to previously unseen instructions.