Evaluating Frontier LLMs on PhD-Level Mathematical Reasoning: A Benchmark on a Textbook in Theoretical Computer Science about Randomized Algorithms

TL;DR

This paper benchmarks four advanced large language models on graduate-level mathematical reasoning tasks from a textbook on randomized algorithms, assessing their accuracy, logical coherence, and reliability in formal proof generation.

Contribution

It provides the first comprehensive evaluation of frontier LLMs on a canonical graduate-level mathematics curriculum, highlighting their strengths and limitations in formal reasoning.

Findings

Top-tier models achieve approximately 66% accuracy in proof generation.

Models show significant variance in consistency and logical coherence.

Frontier models are promising for pedagogical use but need improvements for rigorous proof tasks.

Abstract

The rapid advancement of large language models (LLMs) has led to significant breakthroughs in automated mathematical reasoning and scientific discovery. Georgiev, G${\'o}$mez-Serrano, Tao, and Wagner [GGSTW+25] demonstrate that AI systems can explore new constructions and improve existing bounds, illustrating the growing potential of LLMs to accelerate mathematical discovery. Similarly, Bubeck et al. [BCE+25] show that GPT-5 can meaningfully contribute to scientific workflows, from proposing hypotheses to generating proofs and analyses. Despite these advances, a rigorous evaluation of these models on canonical, graduate-level mathematical theory remains necessary to understand their baseline reasoning capabilities. In this paper, we present a comprehensive benchmark of four frontier models: GPT-5-Thinking, Gemini-3-Pro, Claude-Sonnet-4.5-Thinking, and Grok-4 against the classic curriculum of Randomized Algorithms by Motwani and Raghavan [MR95]. We tasked each model with generating formal LaTeX proofs for a series of lemmas and exercises spanning the textbook. We find that while the top-tier models (Gemini, and Claude) achieve a high accuracy rate (approx. 66%), demonstrating a robust grasp of probabilistic method and formal logic, other models lag significantly in consistency (approx. 40%). We provide a qualitative analysis of the generated proofs, highlighting differences in conciseness, hallucination rates, and logical structure. Our results suggest that while frontier models have reached a threshold of proficiency suitable for graduate-level pedagogical assistance and formalization, significant variance exists in their reliability for rigorous mathematical derivation. The code and the full set of LLM-generated responses are open-sourced and publicly available at https://github.com/magiclinux/math_benchmark_probability.