Fig. 2From: Conditional reduction of the loss value versus reinforcement learning for biassing a de-novo drug design generatorlogP distribution for molecules generated by G versus RL-biased-G and CRLV-biased-G respectivelyBack to article page