Sign up to save your podcasts
Or
Share EP181: Small models beating GPT-5 with logic
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
EP181: Small models beating GPT-5 with logic
Title: SGA-MCTS: Decoupling Planning from Execution via Training-Free Atomic Experience Retrieval
Source: http://arxiv.org/abs/2604.14712v1
Summary:
This work presents a novel framework that amortizes the high cost of inference-time search by casting LLM planning as non-parametric retrieval of symbolic 'SGA atoms.' By enabling System 2 reasoning depth at System 1 speeds without task-specific fine-tuning, it establishes a new efficiency-reasoning Pareto frontier for agentic planning.
View all episodes
By Yun WuTitle: SGA-MCTS: Decoupling Planning from Execution via Training-Free Atomic Experience Retrieval
Source: http://arxiv.org/abs/2604.14712v1
Summary:
This work presents a novel framework that amortizes the high cost of inference-time search by casting LLM planning as non-parametric retrieval of symbolic 'SGA atoms.' By enabling System 2 reasoning depth at System 1 speeds without task-specific fine-tuning, it establishes a new efficiency-reasoning Pareto frontier for agentic planning.