31 March 2025 Monday
| setup | |
| --- | ------------------ |
| 9 | teaching |
| Teaching | |
| 10 | setup |
| setup | |
| 11 | Synthetic Datasets |
| Synthetic Datasets | |
| 12 | Hoops |
| Hoops | |
| 1 | Hoops |
| Hoops | |
| 2 | Synthetic Datasets |
| Synthetic Datasets | |
| 3 | To Pitt |
| Colloq | |
| 4 | Colloq |
| Walk to Cyert |
W: 5 DW: 3 OT: 2 ex
—
The ONE thing:
[ ]Synthetic Data Variance Study
—
To Do
- Mail french press
- Synthetic Data Variance Study
- Setup April
- [>>] Review March
[ ]Call Liberty mutual re:tree down[ ]Understand Chuyuan's slides[ ]Hemisphere plots- [>>] Write up bootstraping study
- [>>] Schedule FNAL badging appointment
[ ]Make sum/SB/SR plots presentation- Move on Jim Parard
- [>>] HGC Slides
- [>>] Update the ETC docs
- [>>] Taxes
- [>>] Turnaround report
- [>>] Understand changes in CMU CPI / UCSB CPI
—
Up at 7
M&T to school … H with me
Moving forward on Jim
Taught
- Probability and Wave function
- Chance found my mug !
setup
- print french press return label
- draft mail for Jim Picard
- mail to Marina
- Mail to Seth Eisenberg re:investing
Run3 Synthetic Data
- see above
[BROKEN LINK: 9058CACF-33CB-48C6-AD39-A423366C0D23]
- See above
Hoops / Steamroom
- 4s and 5s (2nd court)
- Good passing, good D, nice breaks, and some shots
- 10' at Tmax = 124
Manfred get Lazarus Award !!!
- put your people first
- Really Enjoyed the feeling of someone else winning ! (Girard: "easy when not a rival" )
More Bootstrap study
Sent Picard offer letter
Colloq
- HEP for AI Yoni Kahn
- Theorist : known for methods to detect light DM
- Hard Problem 1: How do we understand training?
- Hard Problem 2: Where do scaling laws come from ?
- Hard Problem 3: How to quantify uncertianty
- Claim: We are very far from how learning works in the brain
- Going to Ensemlble over models
- Learning as a phase transition
- Deeper is better up to a point … Surely f(data, problem) ?
- Average at best
Walk to Cyert… Walk home
Some [BROKEN LINK: Myth of the Rational Voter]
Wes' FNAL Talk
- see above