PulseAugur
中
实时 04:36:19

AI Process, Not Just Output, Key to Human-Machine Distinction, Study Finds

作者 PulseAugur 编辑部 · [2 个来源] ·

A new research paper proposes that analyzing the cognitive processes, rather than just the outputs, is more effective for distinguishing humans from advanced AI agents. The study introduces CogCAPTCHA30, a set of 30 cognitive tasks designed to reveal process-level differences, achieving an 0.88 AUC in distinguishing humans from AI. The research evaluated frontier agents like Claude Sonnet 4.5, GPT-5, and Gemini 2.5 Pro, finding that while fine-tuning on human decisions improves process mimicry, process specification remains a bottleneck for achieving truly human-like cognitive processes. AI

影响 Suggests a new paradigm for AI safety and alignment research, moving beyond output-based evaluations to process-based analysis.

排序理由 Academic paper proposing a new method for distinguishing humans from AI by analyzing cognitive processes.

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

AI Process, Not Just Output, Key to Human-Machine Distinction, Study Finds

报道来源 [2]

  1. arXiv cs.AI TIER_1 English(EN) · Milena Rmus, Mathew D. Hardy, Thomas L. Griffiths, Mayank Agrawal ·

    Process Matters more than Output for Distinguishing Humans from Machines

    arXiv:2605.06524v1 Announce Type: new Abstract: Reliable human-machine discrimination is becoming increasingly important as large language models and autonomous agents are deployed in online settings. Existing approaches evaluate whether a system can produce behavior or responses…

  2. arXiv cs.AI TIER_1 English(EN) · Mayank Agrawal ·

    Process Matters more than Output for Distinguishing Humans from Machines

    Reliable human-machine discrimination is becoming increasingly important as large language models and autonomous agents are deployed in online settings. Existing approaches evaluate whether a system can produce behavior or responses indistinguishable from those of a human, follow…

相关实体

相关话题