Prompting a model that only speaks #DNA, like #Evo2, is tricky. In-context learning proved effective, but its boundaries and inner mechanism are not fully clear. Our preprint focuses on those questions, using natural and synthetic sequences, and tools from #mechint.
www.biorxiv.org/content/10.6...
biorxiv.org
Evo 2 as a classification machine: evidence from in-context learning and mechanistic interpretability
In-context learning (iCL) is an emergent capability of Large Language Models (LLMs), allowing them to perform new tasks at inference time using prompt-injected examples. While extensively studied in L...
