Natural Language Edge Labelling: Decoupling Intent from Execution in Structured LM Reasoning
This addresses the need for more controllable and auditable LM inference, though it appears incremental as an overlay on existing methods like Tree-of-Thoughts.
The paper tackles the problem of controllers for structured LM reasoning entangling intent and execution, leading to brittle and inefficient behavior, by introducing Natural Language Edge Labelling (NLEL), which decouples these aspects and shows anticipated accuracy gains and improved compute efficiency in evaluations on benchmarks like GSM8K and MATH.
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 coarse global knobs and yielding brittle, compute-inefficient, and hard-to-audit behavior. We introduce Natural Language Edge Labelling (NLEL), a labeller-tuner overlay that attaches a free-form natural-language directive to each search edge and translates it into a schema-bounded control vector for decoding, search (branch quotas, exploration $β$), generation bundle size, retrieval mixtures, and verification passes. A labeller $Λ$ emits labels from the parent state and a compact context; a tuner $Ψ$ maps $(P, L, C)\to Π$, with strict schema validation and trust-region projection around safe defaults. Downstream selection remains ToT-style with score $S=μ+βσ$ and depth-annealed $β$. We show NLEL strictly generalizes CoT/ToT, prove an anytime-monotonicity property for top-$k$ selection under label-conditioned bundles, and bound selector shortfall by control-vector distortion, providing decision-relevant justification for guards like trust regions and verification passes. We instantiate $Ψ$ as a prompt-only JSON Parameter Emitter and preregister an evaluation on GSM8K, MATH (subset), StrategyQA, and ARC-Challenge with compute-aware reporting (success@compute, tokens-per-success) and ablations over $Λ$, $Ψ$, trust-region radius, and control quantization; preregistered forecasts anticipate accuracy gains at comparable token budgets and improved success@compute under constraints. NLEL offers an interpretable, model-agnostic interface that separates intent from execution for controllable, auditable LM inference.