NNSight and NDIF

anchor

Findings

• IC-302Llama-3.1 models perform between unigram-inference and bigram-inference on Markov chain ICL, with performance improving monotonically with model scale [eval]
• IC-303Explicitly stating the Markovian structure in the prompt significantly improves Llama-3.1-70B's next-state prediction on the Markov chain task [eval]