AI Trends

Trending story

Gemini 4 Hydrogen Scores Higher at Medium and High Tiers on ARC-AGI-2

1 reports1 story updatesUpdated Nov 21 00:57

Understand this story

AI overview

ARC-AGI-2 results released by ARC Prize show that Gemini 4 Hydrogen scored 89.2% and 90.0% at the medium and high reasoning tiers, respectively, higher than Gemini 3.8 Flash's corresponding scores of 82.9% and 89.2%. However, Gemini 4 Hydrogen scored 65.8% at the low tier, below Gemini 3.8 Flash's 77.5%. In addition, Gemini 4 Hydrogen has an xhigh reasoning tier that Gemini 3.8 Flash does not offer. There are currently no other subsequent developments supported by qualified factual evidence.

AI-generated from public reports

Latest development · Report

Gemini 4 Hydrogen Shows Diverging Results Across Reasoning Levels

Developments

1 confirmed developments

  1. Gemini 4 Hydrogen scores higher on the medium and high tiers of ARC-AGI-2

    Source:ARC Prize

Reports and discussion timeline

Browse reports and discussion by time. Their evidence levels remain distinct.

Nov 21

  1. ARC PrizeFirst-party

    Gemini 4 Hydrogen Shows Diverging Results Across Reasoning Levels

    On ARC-AGI-2, Gemini 4 Hydrogen outperformed Gemini 3.8 Flash at the medium and high reasoning levels, scoring 89.2% and 90.0%, respectively, versus 82.9% and 89.2%. However, at the low level, it trailed the latter with 65.8% versus 77.5%. Gemini 4 Hydrogen also adds an xhigh reasoning level not available in Gemini 3.8 Flash.

Related stories

  1. Related storiesGemini 4 Hydrogen Matches or Outperforms Gemini 3.8 Flash on Two ARC-AGI Benchmarks
  2. Related storiesGemini 4 Hydrogen Matches or Surpasses Its Predecessor’s Scores on Two ARC-AGI Benchmarks
  3. Related storiesARC Prize Encounters API Issues While Testing Gemini 4 Hydrogen
  4. Related storiesGemini 4 Hydrogen Achieves Verified ARC-AGI Scores