Learn/Design a binder
What the lab says
Phi Score
- Trained on
- Phi's self-improving data loop
- Read as
- A ranking. A higher score means a design is more likely to bind than a lower one; it is not a probability.
The loop, on the example campaign
Dyno Psiapical set18 / 5018 of 50 passed. 25 lost at a gate, 2 no result, 3 running, 2 queued. 10 of 18 confirmed in the lab.
Dyno Psiapical domain4 / 504 of 50 passed. 34 lost at a gate, 3 no result, 1 failed job, 4 running, 4 queued. 3 of 4 confirmed in the lab.
RFantibodyapical set6 / 506 of 50 passed. 34 lost at a gate, 2 no result, 3 failed job, 4 running, 1 queued. 2 of 6 confirmed in the lab.
RFantibodyapical domain0 / 500 of 50 passed. 34 lost at a gate, 4 no result, 2 failed job, 5 running, 5 queued. 0 of 0 confirmed in the lab.
Germinalapical set11 / 5011 of 50 passed. 24 lost at a gate, 3 no result, 4 failed job, 8 running. 5 of 11 confirmed in the lab.
Germinalapical domain1 / 501 of 50 passed. 38 lost at a gate, 1 no result, 6 running, 4 queued. 0 of 1 confirmed in the lab.
Passed 40Lost at a gate 189No result 15Failed job 10Running 30Queued 16Bound 20Did not bind 20
Passed designs go to the lab; bound and did not bind come back and recalibrate the thresholds and the score. Illustrative campaign.