Despite not being trained to, it turns out the Pearson correlation between a models AA Intelligence Index score and its ability to generate Base64 encoded responses is 0.91
英文标题：Despite not being trained to, it turns out the Pearson correlation between a models AA Intelligence Index score and its ability to generate Base64 encoded responses is 0.91
I built Encode Bench, an open benchmark that asks a model to solve a task and return the answer as a Base64 payload. The initial result surprised me: across the eight models with matching data in the current nine-model snapshot, Encode Bench pass rate has a Pearson correlation of 0.91 with the
中文
I built Encode Bench, an open benchmark that asks a model to solve a task and return the answer as a Base64 payload. The initial result surprised me: across the eight models with matching data in the current nine-model snapshot, Encode Bench pass rate has a Pearson correlation of 0.91 with the
中文全文翻译正在补齐，可先查看下方英文原文。
English / 原文
I built Encode Bench, an open benchmark that asks a model to solve a task and return the answer as a Base64 payload. The initial result surprised me: across the eight models with matching data in the current nine-model snapshot, Encode Bench pass rate has a Pearson correlation of 0.91 with the