Tasks
Someday Maybe Humer argument for why science and reason assume free will:
Someday Maybe Better Coding with Cursor consume
Someday Maybe Make klX work interactively in iPlot 4b
Someday Maybe Check out Aider and Aider.el emacs
Someday Maybe Read HGC TDR HGC
Someday Maybe Workstudy Lab hand
This is the info I received about the steps of hiring a student worker …
Manfred.
-----–— Forwarded message ----–— From: Krystalyn Kovacs <krystalyn@cmu.edu> Date: Fri, Oct 11, 2024 at 12:57 PM Subject: Student employment information To: Manfred Paulini <paulini@cmu.edu>
- Contact student employment with the job description for approval (career@andrew.cmu.edu) (but they ended up forwarding my email to adegeorg@andrew.cmu.edu) (also not sure if this step is technically necessary, but I did it and they were very helpful)
- Find an approved wage –> https://www.cmu.edu/my-workday-toolkit/quick-guides/restricted/student-pay-ranges.pdf
- Make handshake post (contact career@andrew.cmu.edu to get an employer account set up)
- Collect applications
- Interview students
- Email michelle waters with their employment information –> sets them up on workday
- Email the offer letter to the student, and CC michelle waters. include start date
- If the student is working with student data, they need to complete this: https://www.cmu.edu/career/documents/on-campus-employment/for-supervisors/FERPA-Statement-of-Understanding.pdf
Extra resources: https://www.cmu.edu/career/student-employment/for-supervisors/index.html https://www.cmu.edu/my-workday-toolkit/quick-guides/restricted/undergraduate-student-job-profiles.pdf
Someday Maybe Read [BROKEN LINK: Run2 X->HY->4b] 4b
Someday Maybe Read [BROKEN LINK: Evolution of Silicon Sensor Technology in Particle Physics] !!! HGC
Someday Maybe Why Synthetic selJets better the canJets 4b question
They are not better than tag jets
Someday Maybe synthetic data modelling SvB near 0
Someday Maybe Try Signal Reclustering
Planned Send Oscher Mail
Someday Maybe Think of 1D smearing toy
DONE Setup Plain Text Accounting in Emacs
DONE Setup Finances
DONE Outreahc Action items
Action Items: John to follow-up on bulk clearances events (w/Pizza) John to follow-up on status of SAMs for summer ‘26 Need volunteers SciTech early ‘26 Diana to follup-up with Dr. Benmokhtar on speakers series John to follow-up with Murman re:headsets John to follow-up on website/email list ownership Next meeting is planned for early Oct.
Someday Maybe Check out cursor CI interface
DONE Stretching protocol
Now Start Writing up Higgs Lectures
TODO Setup Auto Mortgage Payement with CFSBank
TODO Come back to Jet clustering: re:coordinate changes and mass
Someday Maybe Add FvT on the fly Evaluation … and Friend Trees
CANCEL Signal Injection with mixed data
DONE Implement Evan SvB
Someday Maybe Finalize FSR Truth Tagging
DONE Move workflows to snakemake.
eg: make a fresh script (Snakemake?) for each job.
[ ]How to do testing vs full ?[ ]How to chage optinos runSvB vs no
Someday Maybe Signal Injection Study for mixing
TODO How to Archive Mail.org ?
Someday Maybe issue-bot
Someday Maybe temp file locations during eval
4b code base I think we need to come back to this once we have good testing cI in place. archive/session-summaries/root-tmp-dir-plan-2026-03-14.md
CANCEL Mail re:fixed samples
Someday Maybe Add More tests with Claude
[ ]Run3 TTbar
Someday Maybe Look into Autoresearch
CANCEL Mail re:outreach meeting
TODO Clean up CLAUD.md fies
DONE Condor monitor
DONE Add Run3 TTbar CI
DONE Read Marginal Revolution Book
DONE SvB Friend Trees for mixedall
DONE More GPTEL presets
DONE Check with Jeff on raises
DONE Review March
DONE Setup April
DONE Look into price theory
DONE Reinbursements
DONE Write-up of cmaster
DONE Writeup of wiki
DONE Finish Update Crypto TXs
DONE Clean lab coats
DONE Ingest Dwarakesh podcasts
DONE Tax followup
DONE MvD Slides
DONE Lab Tech Ads Out
DONE SvB Feynet Friend Trees
DONE Ping Cooper
DONE Survey for Rachel
DONE Do ProgressReport 10 April 2026
DONE PourOver: year being ignored?
TODO Write up Thoughts on AI
DONE 3H replies
TODO Compare SvB and FeynNet
DONE Revert runner envirornment var
DONE Pay taxes
DONE Fix 2d ratios in pourover
DONE VARs
DONE Send Urbanic AN info
DONE Follow up with SAMs
DONE Training for Nominal Run3
DONE Module Summary
CANCEL Follow up with Sams
DONE Mail charger
DONE Add inspection to AirBNB listing
DONE Fix AirBNB door code
DONE NaNs in FvT
DONE Mail to Chris re:September Talk
DONE SvBFeynNet per year
DONE Different Vars from different files
Someday Maybe Check out Grapfana
DONE Fix AirBNB login code message
DONE Go through FBC slides
DONE Write up MAC Meeting 22 April 2026
DONE Write up Mans slides
DONE Send mixed data with instructions for MvD weights
DONE Add 3m1t selection
DONE Get dates / room for FNAL
TODO Write up Thoughts on Institutialism
DONE Check out weird event
The run number is 359998 and event number 1639099603 inside picoAODmixedv0.chunk5.root.
DONE Chat MP re:Cola length of stay for Sindhu
DONE Letter to Shrinidhi
[ ]Need start date
TODO CAM referesher training
DONE Find FNAL Badge
DONE Remove same hemi events when mixing
CANCEL Letter for Jeff
DONE Add vars for mixing
DONE Ping Roy re Leo Stephenson
DONE command in to rerun kinitFNAL from cworkers
DONE Update Finances
DONE Setup May
DONE Review April
DONE Should add memory in the graphana monitor
cluster contention. Memory is the bottleneck:
- rogue01: 245GB / 256GB allocated (96%) — only 1 of your jobs there
- rogue02: 243GB / 256GB allocated (95%) — 6 of your jobs there
CPU is fine (~28% used). Other users are eating most of the memory. Each of your jobs requests 16GB (memmb=16000), and there's only ~10-15GB free per node, so the rest queue with reason (Resources)/(Priority).
Three options:
- Wait. As other tenants finish, your pending jobs will dispatch automatically. Snakemake handles this — no action needed.
- Reduce per-job memory request to fit more in the gaps. The histogram config has workermemory: 5GB × 4 workers = 20GB peak, so dropping memmb below ~16000 risks OOM kills. Probably not worth it.
- Be patient on this batch, then maybe lower tasks: 2 → tasks: 1 in the falcon profile for next runs. Each job currently uses 8 CPUs (2 tasks × 4 cpuspertask) but the runner only spins up 4 Dask workers — so 4
CPUs are idle per job. Halving CPU footprint won't help memory though, and CPU isn't your bottleneck either, so no immediate win.
Bottom line: leave it. Snakemake will dispatch the rest as memory frees.
DONE AE re:sub jobs on falcon
See falcon:~/workers/falcon-console/falcon-dask-proposal.md
DONE Book FNAL Hotel
DONE Debug FeynNet
DONE New mixedall sample… in snakemake (Allow rank 0 and 1)
- Verrify same hemievents are gone
DONE Snakemake New MvD
DONE Comparare Rank 0 and rank 3, 7 neighbors
DONE Add Signal to the Run and MvD processing
DONE Update MvD presentaiont with 3T1M
DONE Progress Report
DONE tmux tool ?
CANCEL buy Barnett Tix
DONE Redo SvB Training in Run3
DONE Compare new and "v0" Mixed datasets
DONE would be good to allow mouse scroll on top of tmux output
TODO Investigate mixed data v4j/additional jet descrepancies
TODO ZZ and ZH Run3 Skims
DONE Review Anikets slides
DONE Mail to Li Zhennan
DONE Need to get ultrasound
DONE Autmoate Finance Numbers
DONE SAMs email
DONE Teachers Program Email
TODO Cehck out http://www.incompleteideas.net/
DONE VARs
DONE Slides for making mixed data public
DONE Finialize SAMs Meeting
TODO Update my website with claude
Examples: https://joshspeagle.com/
DONE Letter for Pritam
DONE Ping re:making mixed data slides public
DONE Send Mortage Check
CANCEL Dr Appoointment ?
DONE Start Tenure Packet
DONE Work index
DONE Chuyuan temporary job
DONE AE for Hbb
DONE Setup June
DONE Review May
DONE May Accounting
DONE Ping Rienhard re:Teachers Program
DONE Mail re clustering to Lei
DONE New Mixed samples for Luka
Waiting Next steps for open data
DONE Write up MAC meeting
DONE Refactor based on ak2
DONE Look into fzf
DONE Look into port forwarding
DONE Look into Tmux commands
DONE Get Gaurdian Bike for M
DONE Inclusive Mixed vs Signal Training
DONE VAR
DONE Finances
DONE MAC Summary
DONE Do SvB Study
DONE Update the classifier inputs
DONE Mail Mortgage
DONE Plots for rank3 and 7 with MvD
DONE SvB Study function of 3b training data …
DONE Review Goals
CANCEL Add the other vars to the histogrmas
DONE Add plots for other candVars
CANCEL Turnaround Report
DONE Push classifier changes
DONE SvB Slides
DONE Update Estimate to Complete
DONE HGC Slides
DONE Venmo Sean
DONE Setup July
DONE Review June
DONE Claude Reinbusements
DONE Verrify new SvB plots and canJet features
DONE Mail Stout
DONE Buy Bob Dylan Albums
DONE REad AEs low pt slides
DONE Check in with Chuyuan
DONE Progress Report
DONE Post PGSS HW
DONE Agusta Rule ?
DONE File for reassessment
TODO Get Assesment papers
DONE Synthetic Dataset slides
TODO Update JetDeclustering with lookup ala Evan
DONE Update airbnb with highchair and packand play
DONE Finances
DONE Review Stats Paper
CANCEL Ping Rachel about letter details
- What is the bar for promotion, ?
- Anything special for going up early.
- Assume the point of the letter is to speak to this and provide the candidates feedback
DONE FNAL Badge
DONE JetDeclustering Workflow
DONE Org-workflow for blogs
DONE Mail re:visit
DONE Process Valentinas Notes
TODO Get Claude to use the CAM Tools
DONE Fermilab Reciepts
DONE Start August
DONE Review July
DONE FIx year as list
Now Broken links in github webpages.. attachment dir moved ?
TODO Plan ttHbb all hadronic
The ONE thing Slides showing SvB Modelling
TODO Setup HH4b production tool
TODO Setup emu skims
Now Read Silicon book
TODO Limits for SvB Study
TODO Plan for MPL Fall 2026