Can language models learn analogical reasoning? Investigating training objectives and comparisons to human performance

Open in new window