Map Prediction and Generative Entropy for Multi-Agent Exploration
Alexander Spinos, Bradley Woosley, Justin Rokisky, Christopher Korpela, John G. Rogers III, Brian A. Bittner · 2025
Read Map Prediction and Generative Entropy for Multi-Agent Exploration by Alexander Spinos, Bradley Woosley, Justin Rokisky, Christopher Korpela, John G. Rogers III, Brian A. Bittner (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-13189.txt.
Description
Alexander Spinos and co-authors give, in a 2025 paper, a team of exploring robots a way to imagine the rooms they have not entered. Robot scouts act only on what their sensors have already mapped. This team's map predictor paints plausible completions of the unknown space, using a fine-tuned image-generation model that filled in simulated city blocks quickly and coherently. Where the model's guesses disagree most, a measure the authors call generative entropy runs high.
The trick is where to send the robots next. Instead of chasing the spots promising the most new information, the team sent its three simulated vehicles toward high-entropy spots, the places the model could not make up its mind about. The imagined map converged on the true scene significantly faster that way. The authors frame uncertainty about the unseen, not just ignorance of it, as the compass for exploration.
Catalog record
| CREATOR | Alexander Spinos, Bradley Woosley, Justin Rokisky, Christopher Korpela, John G. Rogers III, Brian A. Bittner |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-13189