Right Answer, Wrong Method: Shortcut Hacking Misleads the Evaluation of LLM Reasoning on Frontier Science Benchmarks

arXiv:2608.02442v1 Announce Type: cross Abstract: Scientific reasoning benchmarks typically evaluate large language models (LLMs) using final-answer accuracy. However, a correct answer does not necessarily demonstrate the reasoning capability targeted by the problem. We identify Solution Hacking, a failure mode in which an LLM reaches the correct answer through invalid shortcuts, such as numerical search, enumeration, guessing, or answer-first verification, without providing a valid task-targeted derivation. We systematically analyze this phenomenon across difficulty levels, scientific domains, and frontier models. Solution hacking increases sharply with benchmark difficulty, from 2.2% on common problems to 28.3% on Olympiad-level problems and 37.4% on HLE. Moreover, 8.2%-44.1% of answers credited as correct across frontier models are identified as hacked solutions. We further develop expert-inspired anti-hacking strategies, including an automatic judge and a test-time instruction. The results show that suppressing shortcut behavior substantially reduces reported accuracy while having a smaller effect on correct and non-hacked accuracy. These findings reveal that answer-only evaluation can overestimate the scientific reasoning capabilities of frontier LLMs.

Related

• When LLMs Stop Following Steps: A Diagnostic Study of Procedural Execution in Language Models
• Relative Kinetic Utility for Reasoning-Aware Structural Pruning in Large Language Models
• Knowing When to Quit: A Principled Framework for Dynamic Abstention in LLM Reasoning
• MTR-Bench: A Comprehensive Benchmark for Multi-Turn Reasoning Evaluation
• LLMs as Assessors: Right for the Right Reason?

Source: arXiv cs.CL | 2026-08-04