AB-MCTS

Adaptive Branching

ARC-AGI2 from 23 to 30

At each step, a choice is made between "creating a new answer (breadth)" vs "refining an existing answer (depth)", as well as deciding "which LLM (model) to assign the task". This selection is made using

Thompson sampling, which considers performance history to make decisions.

Each node (answer) receives a score (Reward) based on "how well this answer solved the problem"

https://github.com/SakanaAI/treequest for

Monte Carlo Tree Search

Wider or Deeper? Scaling LLM Inference-Time Compute with Adaptive...

Recent advances demonstrate that increasing inference-time computation can significantly boost the reasoning capabilities of large language models (LLMs). Although repeated sampling (i.e.,...

https://arxiv.org/abs/2503.04412

Sakana AI

Inference-Time Scaling and Collective Intelligence for Frontier AI

https://sakana.ai/ab-mcts/

AB-MCTS

Adaptive Branching

ARC-AGI2 from 23 to 30

Recommendations