Developing story · 2 episodes · 15 Oct to 18 Oct
AI Agents Struggle With Accuracy And Language
Recent research highlights persistent limitations in AI systems, specifically regarding multilingual fact correction and general agent performance. The latest findings indicate that top-tier AI agents still struggle to achieve high accuracy, capping out at barely 60% on benchmark tests.
-
Taste-Bench shows top AI agents barely exceed 60% accuracy
Researchers from Microsoft and City University of Hong Kong introduced Taste-Bench, a 502‑question benchmark that evaluates AI agents' ability to make good long‑term decisions during multi‑step…
1 source -
New study finds AI struggles to correct false claims clearly across languages
A collaborative study by researchers from the University of Washington, Allen Institute for AI, LMU Munich, UC Berkeley, Stanford, and the University of Wisconsin-Madison examined how AI models…
2 sources
Who and what
OpenAIGoogleMistral AIQwenGemmaAyaMicrosoftCity University of Hong KongGemini 2.5 ProGPT-5.5Mistral Small 4Qwen 3.5 9B