MAGELLAN: Metacognitive predictions of learning progress guide autotelic LLM agents in large goal spaces
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908994558230528 |
|---|---|
| author | Gaven, Loris Carta, Thomas Romac, Clément Colas, Cédric Lamprier, Sylvain Sigaud, Olivier Oudeyer, Pierre-Yves |
| author_facet | Gaven, Loris Carta, Thomas Romac, Clément Colas, Cédric Lamprier, Sylvain Sigaud, Olivier Oudeyer, Pierre-Yves |
| contents | Open-ended learning agents must efficiently prioritize goals in vast possibility spaces, focusing on those that maximize learning progress (LP). When such autotelic exploration is achieved by LLM agents trained with online RL in high-dimensional and evolving goal spaces, a key challenge for LP prediction is modeling one's own competence, a form of metacognitive monitoring. Traditional approaches either require extensive sampling or rely on brittle expert-defined goal groupings. We introduce MAGELLAN, a metacognitive framework that lets LLM agents learn to predict their competence and LP online. By capturing semantic relationships between goals, MAGELLAN enables sample-efficient LP estimation and dynamic adaptation to evolving goal spaces through generalization. In an interactive learning environment, we show that MAGELLAN improves LP prediction efficiency and goal prioritization, being the only method allowing the agent to fully master a large and evolving goal space. These results demonstrate how augmenting LLM agents with a metacognitive ability for LP predictions can effectively scale curriculum learning to open-ended goal spaces. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_07709 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MAGELLAN: Metacognitive predictions of learning progress guide autotelic LLM agents in large goal spaces Gaven, Loris Carta, Thomas Romac, Clément Colas, Cédric Lamprier, Sylvain Sigaud, Olivier Oudeyer, Pierre-Yves Artificial Intelligence Open-ended learning agents must efficiently prioritize goals in vast possibility spaces, focusing on those that maximize learning progress (LP). When such autotelic exploration is achieved by LLM agents trained with online RL in high-dimensional and evolving goal spaces, a key challenge for LP prediction is modeling one's own competence, a form of metacognitive monitoring. Traditional approaches either require extensive sampling or rely on brittle expert-defined goal groupings. We introduce MAGELLAN, a metacognitive framework that lets LLM agents learn to predict their competence and LP online. By capturing semantic relationships between goals, MAGELLAN enables sample-efficient LP estimation and dynamic adaptation to evolving goal spaces through generalization. In an interactive learning environment, we show that MAGELLAN improves LP prediction efficiency and goal prioritization, being the only method allowing the agent to fully master a large and evolving goal space. These results demonstrate how augmenting LLM agents with a metacognitive ability for LP predictions can effectively scale curriculum learning to open-ended goal spaces. |
| title | MAGELLAN: Metacognitive predictions of learning progress guide autotelic LLM agents in large goal spaces |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2502.07709 |