I found reasoning models are much more faithful at text related tasks too (i.e. 1. translating long key-value pairs (i.e. Localizable.strings), 2. long transcri

I found reasoning models are much more faithful at text related tasks too (i.e. 1. translating long key-value pairs (i.e. Localizable.strings), 2. long transcript fixing and verification; 3. look at csv / tabular data and fix) probably due to the reflection mechanism built into these reasoning models. Using prompts such as "check your output to make sure it covers everything in the input" letting the model to double-check its work, avoiding more manual checks on my end.