Skip to content
conftrace_
2026 AACL AACL 2026

Probing Reasoning Length Control in Language Models via Neurons

AbstractReasoning-oriented language models can generate thousands of tokens before answering, yet the internal factors that determine their reasoning length remain poorly understood. This paper studies whether reasoning length can be controlled through sparse neuron interventions, complementing prompt-based, budget-based, and controller-based approaches. The proposed method discovers reasoning-length-associated neurons from paired long and concise correct trajectories on a training split, ranks MLP neurons by long-minus-concise activation contrast, and tests the selected neurons on disjoint evaluation splits by scaling their activations during generation. The evaluation covers math, science, and commonsense reasoning tasks, including GSM8K, MATH500, ARC-Challenge, CommonsenseQA, and QASC, under a strict split-separated protocol that never discovers neurons on held-out evaluation examples. In several, but not all, task/model/budget regimes, selected neurons modulate generated length differently from layer-matched random controls, with a modest single-run specificity result on CommonsenseQA accompanied by a small accuracy increase. These results provide task-, model-, and prompt-protocol-specific evidence that sparse MLP neurons can, in some split-controlled regimes, provide neuron-level control signals for generated reasoning length.

More from AACL 2026

All AACL 2026 papers →