@wangleiphy : I think this is based on an unfair comparison of "filtering out invalid SMILES" versus "retaining all SELFIES". Filtering out invalid SMILES improves the model's performance because they have low likelihoods. In fact, one can also filter out SELFIES with low likelihoods. • TwiCopy

Lei Wang

@wangleiphy

+ Follow

Computational quantum physics ∩ Machine learning. Researcher at Chinese Academy of Sciences

ID: 2431184037

linkhttps://wangleiphy.github.io/ calendar_today23-03-2014 14:55:17

227 Tweet

963 Followers

1,1K Following

Lei Wang

@wangleiphy

5 months ago

I think this is based on an unfair comparison of "filtering out invalid SMILES" versus "retaining all SELFIES". Filtering out invalid SMILES improves the model's performance because they have low likelihoods. In fact, one can also filter out SELFIES with low likelihoods.

thumb_up_off_alt2

chat_bubble_outline1

repeat0

shareShare