Table 5.

Table for comparing metrics on 1000 generated molecules for each context token sequence

	Token sequence SMILES	Unconditional Uniqueness at 1k [%] / SM [%]	LogP {-2, 0, 2} MAD / Uniqueness at 1k [%] / SM [%]	SAScore {2, 3, 4} MAD / Uniqueness at 1k [%] / SM [%]	Molecular Weight {2, 3, 4} MAD / Uniqueness at 1k [%] / SM [%]
1	c1ccccc1 (Benzene)	99.8 / 96.59	0.4 / 99.9 / 75.08	0.15 / 100.0 / 88.17	0.11 / 99.2 / 93.17
2	s1cccc1 (Thiophene)	94.08 / 70.88	0.5 / 97.36 / 53.05	0.15 / 98.39 / 53.33	0.13 / 95.97 / 60.52
3	CC1=CSC=C1 (3-Methylthiophene)	90.78 / 79.86	0.43 / 95.85 / 53.5	0.14 / 93.05 / 62.94	0.14 / 94.04 / 56.97
4	CCO (Ethanol)	99.9 / 61.83	0.17 / 99.8 / 65.73	0.09 / 100.0 / 65.06	0.07 / 99.9 / 54.92
5	CC=O (Acetaldehyde)	99.9 / 89.1	0.19 / 99.2 / 93.67	0.19 / 95.77 / 88.61	0.08 / 97.38 / 91.03
6	CC(=O)OC1=CC=CC= C1C(=O)O (Aspirin)	56.44 / 96.78	0.56 / 73.63 / 87.55	0.2 / 81.45 / 88.41	0.14 / 45.21 / 64.98
7	CC(=O)NC1=CC=C(C =C1)O (Paracetamol)	89.74 / 65.29	0.31 / 92.63 / 72.98	0.15 / 96.12 / 72.55	0.11 / 70.29 / 82.68
8	CN1C=NC2=C1C(=O)N (C(=O)N2C)C (Caffeine)	42.66 / 98.19	0.7 / 70.53 / 91.93	0.23 / 61.69 / 95.69	0.29 / 51.78 / 68.1
9	CN1CCC23C4C1CC5=C2 C(=C(C=C5)O)OC3C (C=C4)O (Morphine)	14.21 / 99.37	0.5 / 46.32 / 94.66	1.85 / 31.86 / 99.49	0.11 / 29.4 / 96.68
10	OC(=O)C(C)c1ccc(cc 1)CC(C)C (Ibuprofen)	33.2 / 44.5	1.03 / 63.65 / 87.46	0.28 / 48.28 / 69.39	0.12 / 30.4 / 66.97