Failing to Explore: Language Models on Interactive Tasks
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910005402271744 |
|---|---|
| author | JafariRaviz, Mahdi Rezaei, Keivan Moakhar, Arshia Soltani Sodagar, Zahra Cheng, Yize Feizi, Soheil |
| author_facet | JafariRaviz, Mahdi Rezaei, Keivan Moakhar, Arshia Soltani Sodagar, Zahra Cheng, Yize Feizi, Soheil |
| contents | We evaluate language models on their ability to explore interactive environments under a limited interaction budget. We introduce three parametric tasks with controllable exploration difficulty, spanning continuous and discrete environments. Across state-of-the-art models, we find systematic under-exploration and suboptimal solutions, with performance often significantly worse than simple explore--exploit heuristic baselines and scaling weakly as the budget increases. Finally, we study two lightweight interventions: splitting a fixed budget into parallel executions, which surprisingly improves performance despite a no-gain theoretical result for our tasks, and periodically summarizing the interaction history, which preserves key discoveries and further improves exploration. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_22345 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Failing to Explore: Language Models on Interactive Tasks JafariRaviz, Mahdi Rezaei, Keivan Moakhar, Arshia Soltani Sodagar, Zahra Cheng, Yize Feizi, Soheil Machine Learning We evaluate language models on their ability to explore interactive environments under a limited interaction budget. We introduce three parametric tasks with controllable exploration difficulty, spanning continuous and discrete environments. Across state-of-the-art models, we find systematic under-exploration and suboptimal solutions, with performance often significantly worse than simple explore--exploit heuristic baselines and scaling weakly as the budget increases. Finally, we study two lightweight interventions: splitting a fixed budget into parallel executions, which surprisingly improves performance despite a no-gain theoretical result for our tasks, and periodically summarizing the interaction history, which preserves key discoveries and further improves exploration. |
| title | Failing to Explore: Language Models on Interactive Tasks |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2601.22345 |