Researcher Announces Upcoming GeoGuesser RL Open Source Release
Tweet from Adithya S K at Hugging Face promises full reproduction materials for training a VLM on the game.
Adithya S K posted that one can literally apply reinforcement learning to a 4B VLM to ace GeoGuesser. He stated that open-source code, RL environment, dataset, training setup, evals, and all materials needed to reproduce the work end-to-end are dropping soon. The post includes a 34-second screen recording of a dark-themed GeoGuesser RL Environment UI for the Qwen3.5-4B model, showing a street-view panorama on the left side.
Combined views
29.1K
1 post, first seen 11h ago
Researcher Announces Upcoming GeoGuesser RL Open Source Release
Tweet from Adithya S K at Hugging Face promises full reproduction materials for training a VLM on the game.
Adithya S K posted that one can literally apply reinforcement learning to a 4B VLM to ace GeoGuesser. He stated that open-source code, RL environment, dataset, training setup, evals, and all materials needed to reproduce the work end-to-end are dropping soon. The post includes a 34-second screen recording of a dark-themed GeoGuesser RL Environment UI for the Qwen3.5-4B model, showing a street-view panorama on the left side.
Reactions
1 postAS 
Adithya S K@adithya_s_kYou can literally RL a 4B VLM to ace GeoGuesser 🌍 > Open-source code, RL environment, dataset, training setup, evals, and everything you need to reproduce it end-to-end > dropping soon!!
- likes: 547
- replies: 28
- bookmarks: 333
- reposts: 31