Posts

🚀 Excited to announce that our paper, “Speaking Your Language: Spatial Relationships in Interpretable Emergent Communication,” has been accepted to NeurIPS 2024 in Vancouver, CA! 🎉 In this work, we explore how agents develop languages to express spatial relationships in their observations.

I’m excited to make public my latest project, nobelGPT, which is focused on creating a Polish-language toy GPT model. Using the llama 3.1 and GPT-2/3 architectures, I’ve built a smaller model trained on texts by Polish Nobel laureates.