Scaling over Scaling: Exploring Test-Time Scaling Plateau in Large Reasoning Models

TL;DR

This paper investigates the limits of test-time scaling in large reasoning models, deriving theoretical bounds for resource allocation and validating them empirically to optimize reasoning performance.

Contribution

It introduces the TTSPM, analyzes scaling paradigms, and derives saturation points, providing a unified framework for understanding test-time scaling limits.

Findings

Identified saturation thresholds for scaling strategies

Unified theoretical bounds for parallel and sequential scaling

Validated bounds on reasoning benchmarks

Abstract

Large reasoning models (LRMs) have exhibited the capacity of enhancing reasoning performance via internal test-time scaling. Building upon this, a promising direction is to further scale test-time compute to unlock even greater reasoning capabilities. However, as we push these scaling boundaries, systematically understanding the practical limits and achieving optimal resource allocation becomes a critical challenge. In this paper, we investigate the scaling plateau of test-time scaling and introduce the Test-Time Scaling Performance Model (TTSPM). We theoretically analyze two fundamental paradigms for such extended scaling, parallel scaling and sequential scaling, from a probabilistic modeling perspective. Our primary contribution is the derivation of the saturation point on the scaling budget for both strategies, identifying thresholds beyond which additional computation yields diminishing returns. Remarkably, despite their distinct mechanisms, both paradigms converge to a unified mathematical structure in their upper bounds. We empirically validate our theoretical findings on challenging reasoning benchmarks, including AIME, MATH-500, and GPQA, demonstrating the practical utility of these bounds for test-time resource allocation. We hope that this work provides insights into the cost-benefit trade-offs of test-time scaling, guiding the development of more resource-efficient inference strategies for large reasoning models.