Showing cs.AIShow all
2 papers · 1 filter
cs.AI2025
Natural Language Edge Labelling: Decoupling Intent from Execution in Structured LM Reasoning
Abhinav Madahar
Controllers for structured LM reasoning (e.g., Chain-of-Thought, self-consistency, and Tree-of-Thoughts) often entangle what to try next with how to execute it, exposing only coars…
cs.AI2025
Lateral Tree-of-Thoughts Surpasses ToT by Incorporating Logically-Consistent, Low-Utility Candidates
Abhinav Madahar
Modern deployments increasingly allocate large test-time compute (thousands of tokens or many node expansions) to boost reliability. Under such budgets, standard Tree-of-Thoughts-s…