Trending storyDeveloping
Elorian releases Glimpse, its first visual reasoning model, and DecoyCount benchmark results
1 report1 sourceupdated 2 hours ago
What happened
AI digest
On October 29, 2026, Latent Space reported that Elorian released Glimpse, its first visual reasoning model, and disclosed how it performed on DecoyCount, a benchmark built with Mercor. The report presents the benchmark as a test of visual counting and reasoning, and says frontier models score under 50% accuracy on it. Glimpse scored 50.48% in that internal test, the only model above 50%. The disclosed progress covers the model release and these internal test results.
Written by AI from the coverage · updated 2 hours ago
Coverage
Follow the reports to see the story from different sides.
Oct 29
- Latent SpaceCan Frontier Models See Past a Decoy? Elorian Tests Visual Reasoning
Elorian 发布首个视觉推理模型 Glimpse，并在与 Mercor 合作的 DecoyCount 基准内部测试中取得 50.48% 准确率，是唯一超过 50% 的模型。
Heat over time
Not enough continuous observations to draw a trend yet.