3 papers
cs.AI2025
Generative Emergent Communication: Large Language Model is a Collective World Model
Tadahiro Taniguchi, Ryo Ueda, Tomoaki Nakamura +2
Large Language Models (LLMs) have demonstrated a remarkable ability to capture extensive world knowledge, yet how this is achieved without direct sensorimotor experience remains a…
cs.CL2025
Metropolis-Hastings Captioning Game: Knowledge Fusion of Vision Language Models via Decentralized Bayesian Inference
Yuta Matsui, Ryosuke Yamaki, Ryo Ueda +2
We propose the Metropolis-Hastings Captioning Game (MHCG), a method to fuse knowledge of multiple vision-language models (VLMs) by learning from each other. Although existing metho…
cs.CL2025
Syntactic Learnability of Echo State Neural Language Models at Scale
Ryo Ueda, Tatsuki Kuribayashi, Shunsuke Kando +1
What is a neural model with minimum architectural complexity that exhibits reasonable language learning capability? To explore such a simple but sufficient neural language model, w…