Unverified50% confidenceFactExact time
LLMs split text into tokens for processing, and 'strawberry' is often segmented into subwords such as 'straw' and 'berry,' causing models to easily make errors when counting at the letter level.
1
Sources
50%
Confidence
Long-term
Relevance
9/15/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn testing, GLM-5.2 correctly answered that 'strawberry' contains 3 R's, but gave a wrong answer the first time it was asked how many S's are in 'occasion', and corrected it to 2 after enabling DeepThink.65% similarUnverifiedGLM-5.2 answered correctly on the Strawberry character counting test but answered incorrectly on the Occasion character counting test.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/922893API
curl https://kongchang.com/api/v1/knowledge/claims/922893MCP
get_claim(id=922893)