# Interactive LLM-assisted Curriculum Learning for Multi-Task Evolutionary Policy Search

arXiv: https://arxiv.org/abs/2602.10891 · HTML: https://mlsanity.com/paper/2602.10891/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 66 references that are in the map.
- First posted on arXiv 11 Feb 2026.

## Builds on (most cited first, 24 of 24)

- [A Survey on Multi-Task Learning](https://mlsanity.com/paper/1707.08114/) (2017)
- [Curriculum Learning: A Survey](https://mlsanity.com/paper/2101.10382/) (2021)
- [Automated Curriculum Learning for Neural Networks](https://mlsanity.com/paper/1704.03003/) (2017)
- [Teacher-Student Curriculum Learning](https://mlsanity.com/paper/1707.00183/) (2017)
- [Reverse Curriculum Generation for Reinforcement Learning](https://mlsanity.com/paper/1707.05300/) (2017)
- [Multi-Task Reinforcement Learning with Soft Modularization](https://mlsanity.com/paper/2003.13661/) (2020)
- [Multi-task Deep Reinforcement Learning with PopArt](https://mlsanity.com/paper/1809.04474/) (2018)
- [Learning Modular Neural Network Policies for Multi-Task and Multi-Robot Transfer](https://mlsanity.com/paper/1609.07088/) (2016)
- [Concise Thoughts: Impact of Output Length on LLM Reasoning and Cost](https://mlsanity.com/paper/2407.19825/) (2024)
- [OMNI-EPIC: Open-endedness via Models of human Notions of Interestingness with Environments Programmed in Code](https://mlsanity.com/paper/2405.15568/) (2024)
- [Curriculum Learning of Multiple Tasks](https://mlsanity.com/paper/1412.1353/) (2014)
- [MarioGPT: Open-Ended Text2Level Generation through Large Language Models](https://mlsanity.com/paper/2302.05981/) (2023)
- [EnvGen: Generating and Adapting Environments via LLMs for Training Embodied Agents](https://mlsanity.com/paper/2403.12014/) (2024)
- [Adaptive Scheduling for Multi-Task Learning](https://mlsanity.com/paper/1909.06434/) (2019)
- [Quality Diversity for Multi-task Optimization](https://mlsanity.com/paper/2003.04407/) (2020)
- [EduPlanner: LLM-Based Multi-Agent Systems for Customized and Intelligent Instructional Design](https://mlsanity.com/paper/2504.05370/) (2025)
- [Beyond Functional Correctness: Investigating Coding Style Inconsistencies in Large Language Models](https://mlsanity.com/paper/2407.00456/) (2024)
- [On the Statistical Benefits of Curriculum Learning](https://mlsanity.com/paper/2111.07126/) (2021)
- [AI Coders Are Among Us: Rethinking Programming Language Grammar Towards Efficient Code Generation](https://mlsanity.com/paper/2404.16333/) (2024)
- [Let the Model Decide its Curriculum for Multitask Learning](https://mlsanity.com/paper/2205.09898/) (2022)
- [Dominated Novelty Search: Rethinking Local Competition in Quality-Diversity](https://mlsanity.com/paper/2502.00593/) (2025)
- [Learning and Sharing: A Multitask Genetic Programming Approach to Image Feature Learning](https://mlsanity.com/paper/2012.09444/) (2020)
- [LLM-POET: Evolving Complex Environments using Large Language Models](https://mlsanity.com/paper/2406.04663/) (2024)
- [Gegelati: Lightweight Artificial Intelligence through Generic and Evolvable Tangled Program Graphs](https://mlsanity.com/paper/2012.08296/) (2020)

## Built on by (most cited first, 1 of 1)

- [LLM-to-Flow Distillation: Teaching Natural Language Goals to Scientific Generators](https://mlsanity.com/paper/2610.17246/) (2026)

## Abstract (arXiv)

Multi-task policy search is a challenging problem because policies are required to generalize beyond training cases. Curriculum learning has proven to be effective in this setting, as it introduces complexity progressively. However, designing effective curricula is labor-intensive and requires extensive domain expertise. LLM-based curriculum generation has only recently emerged as a potential solution, but was limited to operate in static, offline modes without leveraging real-time feedback from the optimizer. Here we propose an interactive LLM-assisted framework for online curriculum generation, where the LLM adaptively designs training cases based on real-time feedback from the evolutionary optimization process. We investigate how different feedback modalities, ranging from numeric metrics alone to combinations with plots and behavior visualizations, influence the LLM ability to generate meaningful curricula. Through a 2D robot navigation case study, tackled with genetic programming as optimizer, we evaluate our approach against static LLM-generated curricula and expert-designed baselines. We show that interactive curriculum generation outperforms static approaches, with multimodal feedback incorporating both progression plots and behavior visualizations yielding performance competitive with expert-designed curricula. This work contributes to understanding how LLMs can serve as interactive curriculum designers for embodied AI systems, with potential extensions to broader evolutionary robotics applications.