English(EN) Claude Fable 5 Reads a Chart to Under 1%. It Still Can’t Count to 23

Claude Fable-5 读图精度高但计数能力欠佳

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-13 22:01

一位用户通过提供包含已知数值数据的图表来测试 Anthropic 的 Claude Fable-5 模型。该模型在读取这些图表时表现出惊人的精度，能够准确识别到小数点后的数值。然而，测试也揭示了该模型在特定计数任务上的局限性，未能正确识别数字 23。 AI

影响突出了当前大型语言模型能力的具体优势和劣势，为未来的开发和用例提供了信息。

排序理由用户对 AI 模型能力进行的评估。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 16 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 用户对 AI 模型能力进行的评估。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

Medium — Claude tag TIER_1 English(EN) · Pasquale Molinaro · 2026-09-13 22:01

Claude Fable 5 准确率达 1% 以下，但仍无法数到 23

<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://pub.towardsai.net/claude-fable-5-reads-a-chart-to-under-1-it-still-cant-count-to-23-0673ae0c7c4c?source=rss------claude-5"><img src="https://cdn-images-1.medium.com/max/1597/1*MtsxSw2WLosXFzy0GNmxNQ.png" …