AI Achieves a Perfect LSAT Score

TL;DR

This paper demonstrates that a language model can achieve a perfect score on the LSAT, challenging the notion that such reasoning tests are exclusive to humans.

Contribution

It shows that a fine-tuned language model can reach perfect LSAT scores, with insights into the effects of prompting, reasoning, and model distillation.

Findings

Models achieve perfect LSAT scores in experiments.

Ablating the thinking phase reduces accuracy, especially in logical reasoning.

Fine-tuning with QLoRA improves logical reasoning performance.

Abstract

This paper reports the first documented instance of a language model achieving a perfect score on an officially disclosed Law School Admission Test (LSAT). Controlled experiments on eight reasoning models show that varying the prompt, shuffling answer choices, and sampling multiple responses have no meaningful effect as drivers of performance. Ablating the thinking phase that models generate before answering, however, lowers frontier accuracy by up to 8 percentage points, predominantly in logical reasoning. Distilled models produce full thinking traces in the same format yet plateau far below frontier performance. A pilot process reward model fine-tuned via QLoRA on official LSAT explanations narrows this gap through Best-of-5 selection, with gains again predominantly in logical reasoning. The gatekeeper of elite legal education since 1948, the LSAT has not merely been passed but answered without a single error by models that reason. The upper bound of the cognitive capacities it has tested is no longer exclusive to human cognition.