2025/5/30 7:00 - 6/3 23:00 : study the paper of AlphaFold 3: Accurate structure prediction of biomolecular interactions with AlphaFold 3
2025/5/31: {"push ups + Mountain Climbers": 460}, {"cartwheels": 100 meters}, {"running": 5km}
2025/5/31 6:46-7:37: section "Network architecture and training"
2025/5/31 7:37-10:12: section "Accuracy across complex types"
2025/5/31 10:12-13:08: section "Predicted confidences track accuracy"
2025/5/31 13:08-14:29: section "Model limitations"
2025/5/31 14:29-14:57: section "Discussion"
Let’s tackle the hardest part now—this is the core and most challenging section of the paper.
2025/5/31 14:57-: section "Methods"
2025/5/31 14:57-: Full algorithm details, Supplementary information(https://static-content.springer.com/esm/art%3A10.1038%2Fs41586-024-07487-w/MediaObjects/41586_2024_7487_MOESM1_ESM.pdf), 45 pages in total
2025/5/31 14:57-15:50: 1 Notation
2025/5/31 16:30-17:00: run 5km in 28min
2025/5/31 20:27-23:49: 2.1 Parsing(To be continued)
2025/6/1: {"push ups + Mountain Climbers": 110}, {"cartwheels": 0}, {"running": 21km}, {"pull ups + Parallel Bar Dips": 100}
2025/6/1 6:09-8:21: run a half marathon in 1:51:53
2025/6/1 13:20-23:30: 2.1 Parsing
2025/6/2: {"pull ups + Parallel Bar Dips": 300}
2025/6/2 6:45-7:06: 2.1 Parsing
2025/6/2 9:04-16:50 (Very sleepy, low efficiency): 2.2 Genetic search
2025/6/3 6:48-: 2.2 Genetic search
0. Run 10km in 40 minutes, run a half marathon in 1 hour and 30 minutes, do 300 push-ups in a row, do 50 pull-ups in a row
History has shown that many great victories were achieved not by those who kept an escape route open, but by those who burned the bridges behind them, forcing themselves to advance against all odds.
