← Back to People
Yona Gal is a researcher at Google DeepMind.
4
Mentions
4
Articles
4
Outlets
Sep 2025 — Sep 2026
Coverage
Positive
Sentiment
Associated Topics
Companies Covered
Coverage Patterns
How media typically covers Yona Gal
Article Types
- research1
Media Angles
- technical1
Narrative Framing
- breakthrough1
Appears In
Sentiment Distribution
Based on 4 scored articles
Positive 100%
Associated AI Models
Gemini 2.5 Pro1GPT-51
Writes About
Appears Alongside
Articles
Most recent firstArticles Written
Yona Gal as author
SimpleQA Verified: A Reliable Factuality Benchmark to Measure Parametric Knowledge
Google DeepMind, Google ResearchresearchpositiveSep 14, 2025
SimpleQA Verified, an improved factuality benchmark for LLMs, reveals that Gemini 2.5 Pro achieves state-of-the-art F1-score of 55.6, outperforming GPT-5 on measuring parametric knowledge and hallucination mitigation.
“Co-author of the SimpleQA Verified benchmark paper”
LLMsAI SafetyFoundation ModelsNLP
Also Mentioned In
Referenced in coverage
Google's new training trick makes LLMs know what they don't know
arXivpositiveSep 2, 2026
“Co-author at Google on reinforcement learning with metacognitive feedback.”
Empty Shelves or Lost Keys? Recall Is the Bottleneck for Parametric Factuality
Google Research, Technion – Israel Institute of TechnologycautiousAug 16, 2026
“Co-author on Empty Shelves or Lost Keys paper studying recall as bottleneck for factuality in LLMs.”
Hallucinations Undermine Trust; Metacognition is a Way Forward
Google Research, Tel Aviv UniversitycautiousMay 10, 2026
“Co-author of paper on hallucinations and metacognition in AI, proposing faithful uncertainty as an alternative approach to reducing confident errors in language models.”