Skip to main content
Back to catalog

\tTokenNameIdentifier

A Roslyn/C# API identifier preceded by a tab character — a code-corpus artifact that is under-trained in Llama 3.

Raw token

"\tTokenNameIdentifier"

Observed behaviors

  • 01Verified under-trained in the Magikarp repetition task for Llama-3-8B, Llama-3.1-8B and Qwen2.5 models — the models cannot reproduce the token on request.GitHub
Observed or documented on
Llama 3Qwen
Behavior types
Under-trained
Tokenizer
Llama-3 BPE
Discovered by
Sander Land & Max Bartolo