Do you reproduce the results of the paper ''IN-CONTEXT REINFORCEMENT LEARNING WITH ALGORITHM DISTILLATION'' ?

[CongryLi]

I haven't found the official code of AD, but there's some new works based on it such as DPT which the authors have released their code. I'm confused if i missed the AD's code. Could you please provide me some suggestion?

[shunzh]

Sorry I missed this issue. I didn't reproduce the results in their paper. For example, I used A2C instead of A3C because A2C is available in stable-baselines3. I'm actually not aware of any official implementation of algorithm distillation.