3 papers
cs.LG2026
Dynamic Latent Routing
Fangyuan Yu, Xin Su, Amir Abdullah
We investigate the temporal concatenation of sub-policies in Markov Decision Processes (MDP) with time-varying reward functions. We introduce General Dijkstra Search (GDS), and pro…
cs.CL2025
Scaling LLM Pre-training with Vocabulary Curriculum
Fangyuan Yu
Modern language models rely on static vocabularies, fixed before pretraining, in contrast to the adaptive vocabulary acquisition observed in human language learning. To bridge this…
cs.LG2024
Iterative Graph Alignment
Fangyuan Yu, Hardeep Singh Arora, Matt Johnson
By compressing diverse narratives, LLMs go beyond memorization, achieving intelligence by capturing generalizable causal relationships. However, they suffer from local 'representat…