ICLR 2026
The 2026 edition of the International Conference on Learning Representations (ICLR), one of the top academic conferences in the fields of machine learning and deep learning.
Timeline (last 90 days)
Research from ICLR 2026 finds that the accuracy of a single model in multi-turn conversations drops to about 39%.
Research records from ICLR 2026 show that in real fact-checking tasks, the disagreement rate among leading models is as high as 63%.
The BTZSC benchmark evaluated dozens of zero-shot classifiers across 22 datasets, covering NLI models, embedding models, and rerankers, with the associated paper published at ICLR 2026.
EgoDex was accepted into ICLR 2026
All Facts (5)
GEPA (Genetic Pareto Prompt Evolution) is a method proposed in an ICLR 2026 Oral paper.
60%UnverifiedResearch from ICLR 2026 finds that the accuracy of a single model in multi-turn conversations drops to about 39%.
50%UnverifiedResearch records from ICLR 2026 show that in real fact-checking tasks, the disagreement rate among leading models is as high as 63%.
50%UnverifiedThe BTZSC benchmark evaluated dozens of zero-shot classifiers across 22 datasets, covering NLI models, embedding models, and rerankers, with the associated paper published at ICLR 2026.
50%UnverifiedEgoDex was accepted into ICLR 2026
50%