Takuya Akiba

DiffusionBlocks: Block-wise Neural Network Training via Diffusion Interpretation

Reimagining Agent-based Modeling with Large Language Model Agents via Shachi

String Seed of Thought: Prompting LLMs for Distribution-Faithful and Diverse Generation

Extending the Context of Pretrained LLMs by Dropping Their Positional Embedding

Transforming Language Models into Program Interpreters via Execution Trace Chain of Thought

Wider or Deeper? Scaling LLM Inference-Time Compute with Adaptive Branching Tree Search

TAID: Temporally Adaptive Interpolated Distillation for Efficient Knowledge Transfer in Language Models

Drop-Upcycling: Training Sparse Mixture of Experts with Partial Re-initialization

Agent Skill Acquisition for Large Language Models via CycleQD