This is the markdown version of the PyAutoCortex Dashboard, which puts the check-in and each project's resume prompt on your clipboard with a single tap of 📋.
One ledger per science project: what is on the cluster right now, what the human said they would do next, and the last few things they wrote down. Nothing here is inferred from a result — the board hands you the check-in and the resume prompt; what a run meant is yours to log.
Last updated 2026-10-23.
| Where | Count |
|---|---|
| Running | 16 |
| Open | 47 |
| Projects | 12 |
📋 check in since last time
Use the cortex skill and treat this chat as an ongoing place to review scientific projects, discuss results and decide what to investigate next. Read PyAutoCortex/AGENTS.md, its project registry and the relevant project ledgers. Follow project-specific instructions when working within a project.
When I give no particular direction, check in across active projects. On the laptop, use the Cortex pull procedure to retrieve updates through each project’s own sync CLI and report run status. Where that access is unavailable, use the available evidence and state what could not be checked. Follow the Cortex check-in procedure to refresh the board and read back each project’s current position, recent activity, outstanding questions and recorded next steps.
When I name a project, result, question or idea, make that the main focus. Help me recall where we left off, inspect available results, compare measured outputs and retrieve relevant records. Help develop scientific questions or explore explanations only when I ask. Bring in other projects where relevant; do not repeat the full project review on every follow-up.
Present factual results, run status and my previously recorded conclusions. Do not offer scientific interpretations, explanations or hypotheses unless I explicitly ask. When I request interpretation, distinguish evidence from speculation and keep proposed interpretations separate from my accepted conclusions. Record scientific conclusions only when I tell you what to preserve.
Record the observations, conclusions and decisions I ask you to preserve using the Cortex ledger procedures. Keep run records, dated discussion notes and next steps consistent, with links to supporting evidence. Keep scientific records in Cortex and bounded development tasks in Mind.
Help plan follow-up analyses or runs when requested, using the project’s own execution workflow. Submit compute only when I explicitly ask, and preserve the applicable resource and approval requirements. Route implementation changes through the development workflow.
Continue from decisions and authorizations already established in this conversation. After taking action, report what changed, what was recorded and what remains unresolved.
Last check-in: 2026-10-17T11:18Z.
| Project | Running | Open | Last update |
|---|---|---|---|
| subhalo_validation | 0 | 5 | 2026-09-07 |
| subhalo | 0 | 0 | 2026-10-15 |
| euclid_dr1 | 16 | 42 | 2026-10-10 |
| analytic_gaussian | 0 | 0 | 2026-10-23 |
| ep_toy_gaussian | 0 | 0 | 2026-10-14 |
| slope_hierarchy_scale | 0 | 0 | 2026-10-14 |
| ic50_workspace | 0 | 0 | 2026-10-14 |
| autolens_inference | 0 | 0 | 2026-10-21 |
| autofit_inference | 0 | 0 | 2026-10-16 |
| euclid_detailed | 0 | 0 | 2026-10-15 |
| cowls | 0 | 0 | 2026-10-15 |
| cowls_ansofi | 0 | 0 | 2026-10-19 |
subhalo_validation — Do the SLaM subhalo pipelines false-detect on lenses with no subhalo · 0 running · 5 open · updated 2026-09-07
Registered: not recorded
Local: /mnt/c/Users/Jammy/Science/active/subhalo_validation
RAL: /mnt/ral/jnightin/active/subhalo_validation
Current position
Five chains on RAL under the corrected adapt settings: rectangular_adapt on pl_eff_0 / pl_eff_1_outer (job A done, job B 342311 submitted) and the delaunay_adapt_split reruns plus rectangular_adapt on pl_sersic_0. pl_sersic_0 on delaunay_adapt_split is accepted (no false detection, +2.557). Next: pull when the numba chains land and read the evidence_increase of each; no more 1000-pixel runs.
📋
Use the cortex skill. — resume subhalo_validation: read PyAutoCortex projects/subhalo_validation.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/subhalo_validation/wiki/project/state.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (5)
| Job | State | Partition | Date | Description |
|---|---|---|---|---|
342299_1 |
open | ral | 2026-09-07 | delaunay_adapt_split_pl_eff_0: job B, numba chain at the standard 1250-pixel Delaunay mesh (8c / 96gb / 48 h); reloads job A 342231_1; the 09-07 ruling: run al… |
342299_2 |
open | ral | 2026-09-07 | delaunay_adapt_split_pl_eff_1_outer: job B, numba chain at the standard 1250-pixel Delaunay mesh (8c / 96gb / 48 h); reloads job A 342231_2; the 09-07 ruling:… |
342311_1 |
open | ral | 2026-09-07 | rectangular_adapt_pl_eff_0: job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_1; the… |
342311_2 |
open | ral | 2026-09-07 | rectangular_adapt_pl_eff_1_outer: job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_… |
342311_0 |
open | ral | 2026-09-07 | rectangular_adapt_pl_sersic_0: job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_0;… |
Full descriptions (5 shortened above)
342299_1— delaunay_adapt_split_pl_eff_0: job B, numba chain at the standard 1250-pixel Delaunay mesh (8c / 96gb / 48 h); reloads job A 342231_1; the 09-07 ruling: run all datasets at 1250, no more pix1000342299_2— delaunay_adapt_split_pl_eff_1_outer: job B, numba chain at the standard 1250-pixel Delaunay mesh (8c / 96gb / 48 h); reloads job A 342231_2; the 09-07 ruling: run all datasets at 1250, no more pix1000342311_1— rectangular_adapt_pl_eff_0: job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_1; the 09-07 ask: submit the rectangular runs too342311_2— rectangular_adapt_pl_eff_1_outer: job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_2; the 09-07 ask: submit the rectangular runs too342311_0— rectangular_adapt_pl_sersic_0: job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_0; the 09-07 ask: submit the rectangular runs too
Recent log — 5 of 40 entries
- 2026-09-07 — result — accepted delaunay_adapt_split_pl_sersic_0 (R-20260907-01): Accepted. The numba delaunay_adapt_split chain on pl_sersic_0 (job 342273_0, 22 h 08 m, 1250 Hilbert / 1310 Delaunay pixels) lands evidence_increase +2.557 (< 5): no false detection. The best cell parks a 1.6e6 Msun nuisance subhalo at (-1.43, -2.36)", outside the ring. The shift from the legacy -0.30 comes with the S/N-3 adapt cap, adaptive over-sampling and the numba pool, not a code fault. Results are good; run the remaining datasets at 1250 pixels; no more 1000-pixel runs.
- 2026-09-07 — run — 342311_0 submitted: rectangular_adapt_pl_sersic_0 — job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_0; the 09-07 ask: submit the rectangular runs too
- 2026-09-07 — run — 342311_2 submitted: rectangular_adapt_pl_eff_1_outer — job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_2; the 09-07 ask: submit the rectangular runs too
- 2026-09-07 — run — 342311_1 submitted: rectangular_adapt_pl_eff_0 — job B, numba chain (8c / 96gb / 48 h) at the standard 1250-pixel mesh under PIPELINE=rectangular_adapt; reloads job A 342237_1; the 09-07 ask: submit the rectangular runs too
- 2026-09-07 — run — 342299_2 submitted: delaunay_adapt_split_pl_eff_1_outer — job B, numba chain at the standard 1250-pixel Delaunay mesh (8c / 96gb / 48 h); reloads job A 342231_2; the 09-07 ruling: run all datasets at 1250, no more pix1000
active · ral · subhalo_validation#2
subhalo — SLaM subhalo modelling of the SLACS lens sample · no runs recorded · updated 2026-10-15
Registered: 2026-10-15
Local: /mnt/c/Users/Jammy/Science/active/subhalo
RAL: /mnt/ral/jnightin/active/subhalo
Current position
SLACS sample under dataset/, output/ empty, no runs yet; ral_root not yet created on RAL — the first hpc/sync push creates it.
📋
Use the cortex skill. — resume subhalo: read PyAutoCortex projects/subhalo.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/subhalo/README.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
euclid_dr1 — Does the sep1 DR1 delivery reproduce the ten prelim lenses by eye · 16 running · 42 open · updated 2026-10-10
Registered: 2026-09-11
Local: /mnt/c/Users/Jammy/Science/euclid_dr1
RAL: /mnt/ral/jnightin/euclid_dr1
Current position
The top-1000 DR1 sep1 vis_lp array finished 971 of 1000; 29 hit the 18-hour walltime. Array 344645 is now running vis_pix with the Numba CPU sparse route for those 971 completed lenses only.
📋
Use the cortex skill. — resume euclid_dr1: read PyAutoCortex projects/euclid_dr1.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/euclid_dr1/wiki/project/state.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (58)
| Job | State | Partition | Date | Description |
|---|---|---|---|---|
342650 |
open | ral | 2026-09-11 | euclid_dr1 sep1 reproduction: the ten euclid_dr1_prelim lenses refitted from the sep1 delivery through prelim's route (hpc/batch_cpu/submit_initial_lens_model_… |
345279 |
running | ral | 2026-09-20 | VIS-LP JAX CPU pilot for the nine remainder tiles completing the first ten sorted DR1 lenses; single RGB thumbnails uploaded |
345305 |
running | ral | 2026-09-20 | DR1 full vis_lp production batch 1/5: rest_01, 1000 lenses, output root dr1_full |
345669 |
open | ral | 2026-09-20 | DR1 full vis_lp production batch 2/5: rest_02, 1000 lenses, output root dr1_full |
345670 |
open | ral | 2026-09-20 | DR1 full vis_lp production batch 3/5: rest_03, 1000 lenses, output root dr1_full |
345671 |
open | ral | 2026-09-20 | DR1 full vis_lp production batch 4/5: rest_04, 1000 lenses, output root dr1_full |
345672 |
open | ral | 2026-09-20 | DR1 full vis_lp production batch 5/5: rest_05, 1000 lenses, output root dr1_full |
349353 |
open | ral | 2026-09-21 | 100-tile dr1_sep1_rest SED array: Sersic VIS with lens-light n~Uniform(0.5,10.0), then NIR Y/J/H and EXT DECam g/r/i/z; output_sed; max 20 concurrent |
350581 |
open | ral | 2026-09-23 | vis_lp-only inspection bundle over 4,912 completed dr1_sep1_rest tiles (vislp_full_20260923, OUTPUT_DIR=dr1_full, single 24 h CPU job) |
350804 |
running | ral | 2026-09-25 | DR1 dr1_sep1_rest priority-250 vis_lp array (euclid_dr1_vis_lp_prio250), 0-249, manifest hpc/run_manifests/positions_priority_250.txt, positions finder 2.1 pos… |
351085 |
running | ral | 2026-09-25 | DR1 dr1_sep1_rest next-5000 vis_lp batch 1/5 (euclid_dr1_vis_lp_n5k_p01), 1000 lenses, manifest hpc/run_manifests/vis_lp_next5000_20260925_part01.txt |
351086 |
running | ral | 2026-09-25 | DR1 dr1_sep1_rest next-5000 vis_lp batch 2/5 (euclid_dr1_vis_lp_n5k_p02), 1000 lenses, manifest hpc/run_manifests/vis_lp_next5000_20260925_part02.txt |
351087 |
running | ral | 2026-09-25 | DR1 dr1_sep1_rest next-5000 vis_lp batch 3/5 (euclid_dr1_vis_lp_n5k_p03), 1000 lenses, manifest hpc/run_manifests/vis_lp_next5000_20260925_part03.txt |
351088 |
running | ral | 2026-09-25 | DR1 dr1_sep1_rest next-5000 vis_lp batch 4/5 (euclid_dr1_vis_lp_n5k_p04), 1000 lenses, manifest hpc/run_manifests/vis_lp_next5000_20260925_part04.txt |
351089 |
running | ral | 2026-09-25 | DR1 dr1_sep1_rest next-5000 vis_lp batch 5/5 (euclid_dr1_vis_lp_n5k_p05), 1000 lenses, manifest hpc/run_manifests/vis_lp_next5000_20260925_part05.txt |
351090 |
open | ral | 2026-09-25 | DR1 priority-250 vis_pix array (euclid_dr1_vis_pix_prio250), 0-249, dependency afterany:350804 |
351091 |
open | ral | 2026-09-25 | DR1 priority-250 SED array (euclid_dr1_sed_prio250), 0-249 max 60 concurrent, output_sed, dependency afterany:351090 |
356227 |
running | ral | 2026-09-26 | DR1 priority-250 vis_pix by-eye test: tasks 0-9 of submit_initial_lens_model_vis_pix_priority250 (euclid_dr1_vis_pix_prio250_test10), no dependency; full array… |
356319 |
running | ral | 2026-09-26 | DR1 priority-250 SED rerun, tasks 0-9: the ten tiles with no EXT/DECam imaging, VIS + NIR Y/J/H only (euclid_dr1_sed_prio250_nironly); SED scripts no longer re… |
356390 |
open | ral | 2026-09-26 | DR1 next-5000 vis_pix part 01/05 (euclid_dr1_vis_pix_n5k, PART=01): 992 lenses whose 351085 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pi… |
356447 |
open | ral | 2026-09-26 | DR1 next-5000 vis_pix part 02/05 (euclid_dr1_vis_pix_n5k, PART=02): 996 lenses whose 351086 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pi… |
356448 |
open | ral | 2026-09-26 | DR1 next-5000 vis_pix part 03/05 (euclid_dr1_vis_pix_n5k, PART=03): 997 lenses whose 351087 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pi… |
356449 |
open | ral | 2026-09-26 | DR1 next-5000 vis_pix part 04/05 (euclid_dr1_vis_pix_n5k, PART=04): 996 lenses whose 351088 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pi… |
356450 |
open | ral | 2026-09-26 | DR1 next-5000 vis_pix part 05/05 (euclid_dr1_vis_pix_n5k, PART=05): 994 lenses whose 351089 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pi… |
356451 |
open | ral | 2026-09-26 | DR1 next-5000 SED part 01/05 (euclid_dr1_sed_n5k, PART=01): same 992-lens manifest vis_pix_sed_next5000_20260926_part01.txt, vis_lp seed from output/, output_s… |
356452 |
open | ral | 2026-09-26 | DR1 next-5000 SED part 02/05 (euclid_dr1_sed_n5k, PART=02): same 996-lens manifest vis_pix_sed_next5000_20260926_part02.txt, vis_lp seed from output/, output_s… |
356453 |
open | ral | 2026-09-26 | DR1 next-5000 SED part 03/05 (euclid_dr1_sed_n5k, PART=03): same 997-lens manifest vis_pix_sed_next5000_20260926_part03.txt, vis_lp seed from output/, output_s… |
356555 |
open | ral | 2026-09-26 | DR1 next-5000 SED part 04/05 (euclid_dr1_sed_n5k, PART=04): same 996-lens manifest vis_pix_sed_next5000_20260926_part04.txt, vis_lp seed from output/, output_s… |
356556 |
open | ral | 2026-09-26 | DR1 next-5000 SED part 05/05 (euclid_dr1_sed_n5k, PART=05): same 994-lens manifest vis_pix_sed_next5000_20260926_part05.txt, vis_lp seed from output/, output_s… |
374848 |
open | ral | 2026-09-30 | DR1 dr1_sep1_rest v2-grade SED chain batch 01/5 (euclid_dr1_sed_v2g), 998 tiles whose vis_lp 366938 completed, manifest hpc/run_manifests/sed_v2grade_20260930_… |
374849 |
open | ral | 2026-09-30 | DR1 dr1_sep1_rest v2-grade SED chain batch 02/5 (euclid_dr1_sed_v2g), 998 tiles whose vis_lp 367039 completed, manifest hpc/run_manifests/sed_v2grade_20260930_… |
374850 |
open | ral | 2026-09-30 | DR1 dr1_sep1_rest v2-grade SED chain batch 03/5 (euclid_dr1_sed_v2g), 995 tiles whose vis_lp 367141 completed, manifest hpc/run_manifests/sed_v2grade_20260930_… |
374851 |
open | ral | 2026-09-30 | DR1 dr1_sep1_rest v2-grade SED chain batch 04/5 (euclid_dr1_sed_v2g), 995 tiles whose vis_lp 367243 completed, manifest hpc/run_manifests/sed_v2grade_20260930_… |
374852 |
open | ral | 2026-09-30 | DR1 dr1_sep1_rest v2-grade SED chain batch 05/5 (euclid_dr1_sed_v2g), 996 tiles whose vis_lp 367244 completed, manifest hpc/run_manifests/sed_v2grade_20260930_… |
377038 |
open | ral | 2026-09-30 | DR1 rest-batch vis_pix part 02/4 dependent singles (euclid_dr1_vis_pix_rest_p02): 945 single-task jobs ids 377038-377987, afterok:375701_N each (indices 0-944)… |
377988 |
open | ral | 2026-09-30 | DR1 rest-batch vis_pix part 03/4 dependent singles (euclid_dr1_vis_pix_rest_p03): 945 single-task jobs ids 377988-378936, afterok:375702_N each (indices 0-944)… |
378940 |
open | ral | 2026-09-30 | DR1 rest-batch vis_pix part 04/4 dependent singles (euclid_dr1_vis_pix_rest_p04): 944 single-task jobs ids 378940-379889, afterok:375703_N each (indices 0-943)… |
384040 |
running | ral | 2026-10-01 | DR1 dr1_v2_new sep28 eclipse new_ab vis_lp (euclid_dr1_vis_lp_sep28_p01), 987 tiles (942 dr1_v2_new_cutouts_needed + 45 re-centred cut-outs; 48+21 dr1_sep1 dup… |
384046 |
running | ral | 2026-10-01 | DR1 dr1_v2_new sep28 vis_pix dependent singles (euclid_dr1_vis_pix_sep28_p01): 987 single-task jobs ids 384046-385042, each --array=N --dependency=afterok:3840… |
388367 |
open | ral | 2026-10-03 | DR1 missing-SED sweep rest part 01/4 (euclid_dr1_sed_miss_r01): 944 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_… |
388368 |
open | ral | 2026-10-03 | DR1 missing-SED sweep rest part 02/4 (euclid_dr1_sed_miss_r02): 938 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_… |
388369 |
open | ral | 2026-10-03 | DR1 missing-SED sweep rest part 03/4 (euclid_dr1_sed_miss_r03): 943 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_… |
388370 |
open | ral | 2026-10-03 | DR1 missing-SED sweep rest part 04/4 (euclid_dr1_sed_miss_r04): 939 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_… |
388371 |
open | ral | 2026-10-03 | DR1 missing-SED sweep dr1_v2_new (euclid_dr1_sed_miss_v2new): 303 sep28 tiles with completed vis_lp 384040 — first SED on non-DECam ground bands (HSC/MegaCam/P… |
388372 |
open | ral | 2026-10-03 | DR1 missing-SED retry probe (euclid_dr1_sed_miss_retry): tasks 0-49 of 1,252 earlier SED TIMEOUT/FAILED/partial tiles, manifest sed_missing_20261003_retry.txt,… |
398270 |
open | ral | 2026-10-08 | DR1 dr1_sep1 ex-top1000 vis_pix chained on vis_lp 397467 (euclid_dr1_vis_pix_ex_top1000_p01): immediate array 398270 over 256 lenses whose vis_lp had completed… |
400282 |
open | ral | 2026-10-09 | DR1 dr1_sep1 SED fill 3 pass 1 part 01/4 (euclid_dr1_sed_fill3_p01): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move… |
400308 |
open | ral | 2026-10-09 | DR1 dr1_sep1 SED fill 3 pass 1 part 02/4 (euclid_dr1_sed_fill3_p02): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move… |
400309 |
open | ral | 2026-10-09 | DR1 dr1_sep1 SED fill 3 pass 1 part 03/4 (euclid_dr1_sed_fill3_p03): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move… |
400310 |
open | ral | 2026-10-09 | DR1 dr1_sep1 SED fill 3 pass 1 part 04/4 (euclid_dr1_sed_fill3_p04): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move… |
402719 |
running | ral | 2026-10-10 | DR1 dr1_sep1 SED fill 4 pass 1 part 01/4 (euclid_dr1_sed_fill4_p01): indices 697-999 of sed_fill_20261009/part01.txt, replacing 400282's cancelled pending task… |
402720 |
running | ral | 2026-10-10 | DR1 dr1_sep1 SED fill 4 pass 1 part 02/4 (euclid_dr1_sed_fill4_p02): indices 527-999 of sed_fill_20261009/part02.txt, replacing 400308's cancelled pending task… |
402721 |
running | ral | 2026-10-10 | DR1 dr1_sep1 SED fill 4 pass 1 part 03/4 (euclid_dr1_sed_fill4_p03): indices 363-999 of sed_fill_20261009/part03.txt, replacing 400309's cancelled pending task… |
402722 |
running | ral | 2026-10-10 | DR1 dr1_sep1 SED fill 4 pass 1 part 04/4 (euclid_dr1_sed_fill4_p04): indices 803-862 of sed_fill_20261009/part04.txt, replacing 400310's cancelled pending task… |
402909 |
open | ral | 2026-10-10 | DR1 SED fill 4 pass 2 (retry) part 01/4 (euclid_dr1_sed_fill4_retry_p01): indices 0-999 of sed_fill_20261009/part01.txt; same script submit_sersic_waveband_sed… |
402910 |
open | ral | 2026-10-10 | DR1 SED fill 4 pass 2 (retry) part 02/4 (euclid_dr1_sed_fill4_retry_p02): indices 0-999 of sed_fill_20261009/part02.txt; same script submit_sersic_waveband_sed… |
402911 |
open | ral | 2026-10-10 | DR1 SED fill 4 pass 2 (retry) part 03/4 (euclid_dr1_sed_fill4_retry_p03): indices 0-999 of sed_fill_20261009/part03.txt; same script submit_sersic_waveband_sed… |
402912 |
open | ral | 2026-10-10 | DR1 SED fill 4 pass 2 (retry) part 04/4 (euclid_dr1_sed_fill4_retry_p04): indices 0-862 of sed_fill_20261009/part04.txt; same script submit_sersic_waveband_sed… |
Full descriptions (43 shortened above)
342650— euclid_dr1 sep1 reproduction: the ten euclid_dr1_prelim lenses refitted from the sep1 delivery through prelim's route (hpc/batch_cpu/submit_initial_lens_model_two_stage: vis_lp then vis_pix), array 0-9, wall budget 36:00; the by-eye reference is prelim run 342629350804— DR1 dr1_sep1_rest priority-250 vis_lp array (euclid_dr1_vis_lp_prio250), 0-249, manifest hpc/run_manifests/positions_priority_250.txt, positions finder 2.1 positions356227— DR1 priority-250 vis_pix by-eye test: tasks 0-9 of submit_initial_lens_model_vis_pix_priority250 (euclid_dr1_vis_pix_prio250_test10), no dependency; full array 351090 put on scontrol hold until the human has looked356319— DR1 priority-250 SED rerun, tasks 0-9: the ten tiles with no EXT/DECam imaging, VIS + NIR Y/J/H only (euclid_dr1_sed_prio250_nironly); SED scripts no longer refuse no-EXT tiles356390— DR1 next-5000 vis_pix part 01/05 (euclid_dr1_vis_pix_n5k, PART=01): 992 lenses whose 351085 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pix_sed_next5000_20260926_part01.txt, array 0-991%100, 8 CPU/64 GB/18 h, no dependency356447— DR1 next-5000 vis_pix part 02/05 (euclid_dr1_vis_pix_n5k, PART=02): 996 lenses whose 351086 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pix_sed_next5000_20260926_part02.txt, array 0-995%100, 8 CPU/64 GB/18 h, no dependency356448— DR1 next-5000 vis_pix part 03/05 (euclid_dr1_vis_pix_n5k, PART=03): 997 lenses whose 351087 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pix_sed_next5000_20260926_part03.txt, array 0-996%100, 8 CPU/64 GB/18 h, no dependency356449— DR1 next-5000 vis_pix part 04/05 (euclid_dr1_vis_pix_n5k, PART=04): 996 lenses whose 351088 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pix_sed_next5000_20260926_part04.txt, array 0-995%100, 8 CPU/64 GB/18 h, no dependency356450— DR1 next-5000 vis_pix part 05/05 (euclid_dr1_vis_pix_n5k, PART=05): 994 lenses whose 351089 vis_lp log ended Finished vis_lp, manifest hpc/run_manifests/vis_pix_sed_next5000_20260926_part05.txt, array 0-993%100, 8 CPU/64 GB/18 h, no dependency356451— DR1 next-5000 SED part 01/05 (euclid_dr1_sed_n5k, PART=01): same 992-lens manifest vis_pix_sed_next5000_20260926_part01.txt, vis_lp seed from output/, output_sed, array 0-991%60, 6 h, independent of vis_pix; no-EXT tiles fit VIS + NIR only356452— DR1 next-5000 SED part 02/05 (euclid_dr1_sed_n5k, PART=02): same 996-lens manifest vis_pix_sed_next5000_20260926_part02.txt, vis_lp seed from output/, output_sed, array 0-995%60, 6 h, independent of vis_pix; no-EXT tiles fit VIS + NIR only356453— DR1 next-5000 SED part 03/05 (euclid_dr1_sed_n5k, PART=03): same 997-lens manifest vis_pix_sed_next5000_20260926_part03.txt, vis_lp seed from output/, output_sed, array 0-996%60, 6 h, independent of vis_pix; no-EXT tiles fit VIS + NIR only356555— DR1 next-5000 SED part 04/05 (euclid_dr1_sed_n5k, PART=04): same 996-lens manifest vis_pix_sed_next5000_20260926_part04.txt, vis_lp seed from output/, output_sed, array 0-995%60, 6 h, independent of vis_pix; no-EXT tiles fit VIS + NIR only356556— DR1 next-5000 SED part 05/05 (euclid_dr1_sed_n5k, PART=05): same 994-lens manifest vis_pix_sed_next5000_20260926_part05.txt, vis_lp seed from output/, output_sed, array 0-993%60, 6 h, independent of vis_pix; no-EXT tiles fit VIS + NIR only374848— DR1 dr1_sep1_rest v2-grade SED chain batch 01/5 (euclid_dr1_sed_v2g), 998 tiles whose vis_lp 366938 completed, manifest hpc/run_manifests/sed_v2grade_20260930_part01.txt, 8 CPU/16 GB/3 h, %40, partitions ral+gpu; ~18% TIMEOUT expected (Nautilus band hang)374849— DR1 dr1_sep1_rest v2-grade SED chain batch 02/5 (euclid_dr1_sed_v2g), 998 tiles whose vis_lp 367039 completed, manifest hpc/run_manifests/sed_v2grade_20260930_part02.txt, 8 CPU/16 GB/3 h, %40, partitions ral+gpu; ~18% TIMEOUT expected (Nautilus band hang)374850— DR1 dr1_sep1_rest v2-grade SED chain batch 03/5 (euclid_dr1_sed_v2g), 995 tiles whose vis_lp 367141 completed, manifest hpc/run_manifests/sed_v2grade_20260930_part03.txt, 8 CPU/16 GB/3 h, %40, partitions ral+gpu; ~18% TIMEOUT expected (Nautilus band hang)374851— DR1 dr1_sep1_rest v2-grade SED chain batch 04/5 (euclid_dr1_sed_v2g), 995 tiles whose vis_lp 367243 completed, manifest hpc/run_manifests/sed_v2grade_20260930_part04.txt, 8 CPU/16 GB/3 h, %40, partitions ral+gpu; ~18% TIMEOUT expected (Nautilus band hang)374852— DR1 dr1_sep1_rest v2-grade SED chain batch 05/5 (euclid_dr1_sed_v2g), 996 tiles whose vis_lp 367244 completed, manifest hpc/run_manifests/sed_v2grade_20260930_part05.txt, 8 CPU/16 GB/3 h, %40, partitions ral+gpu; ~18% TIMEOUT expected (Nautilus band hang)377038— DR1 rest-batch vis_pix part 02/4 dependent singles (euclid_dr1_vis_pix_rest_p02): 945 single-task jobs ids 377038-377987, afterok:375701_N each (indices 0-944); 8 CPU/16 GB/24 h, partitions ral+gpu; driver submit_vis_pix_rest_chain.sh377988— DR1 rest-batch vis_pix part 03/4 dependent singles (euclid_dr1_vis_pix_rest_p03): 945 single-task jobs ids 377988-378936, afterok:375702_N each (indices 0-944); 8 CPU/16 GB/24 h, partitions ral+gpu; driver submit_vis_pix_rest_chain.sh378940— DR1 rest-batch vis_pix part 04/4 dependent singles (euclid_dr1_vis_pix_rest_p04): 944 single-task jobs ids 378940-379889, afterok:375703_N each (indices 0-943); 8 CPU/16 GB/24 h, partitions ral+gpu; driver submit_vis_pix_rest_chain.sh384040— DR1 dr1_v2_new sep28 eclipse new_ab vis_lp (euclid_dr1_vis_lp_sep28_p01), 987 tiles (942 dr1_v2_new_cutouts_needed + 45 re-centred cut-outs; 48+21 dr1_sep1 duplicates excluded), v2 A 114 / B 873, v2_score 3.000->1.500, manifest hpc/run_manifests/vis_lp_sep28_20261001.txt, 4 CPU/12 GB/18 h, finder-2.1 positions (n_lt_2 279), --array=0-986%100, partition ral only; output/dr1_v2_new384046— DR1 dr1_v2_new sep28 vis_pix dependent singles (euclid_dr1_vis_pix_sep28_p01): 987 single-task jobs ids 384046-385042, each --array=N --dependency=afterok:384040_N --kill-on-invalid-dep=yes (indices 0-986); Numba sparse CPU, 8 CPU/16 GB/24 h, partition ral only; driver hpc/batch_cpu/submit_vis_pix_rest_chain.sh --batch=sep28 (idempotent, re-run to catch stragglers)388367— DR1 missing-SED sweep rest part 01/4 (euclid_dr1_sed_miss_r01): 944 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_missing_20261003_rest_part01.txt, %40, 8 CPU/16 GB/3 h, partition ral only, script hpc/batch_cpu/submit_sersic_waveband_missing (skips complete tiles)388368— DR1 missing-SED sweep rest part 02/4 (euclid_dr1_sed_miss_r02): 938 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_missing_20261003_rest_part02.txt, %40, 8 CPU/16 GB/3 h, partition ral only, script hpc/batch_cpu/submit_sersic_waveband_missing (skips complete tiles)388369— DR1 missing-SED sweep rest part 03/4 (euclid_dr1_sed_miss_r03): 943 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_missing_20261003_rest_part03.txt, %40, 8 CPU/16 GB/3 h, partition ral only, script hpc/batch_cpu/submit_sersic_waveband_missing (skips complete tiles)388370— DR1 missing-SED sweep rest part 04/4 (euclid_dr1_sed_miss_r04): 939 dr1_sep1_rest tiles, vis_lp 375700-375703 done, no SED yet, manifest hpc/run_manifests/sed_missing_20261003_rest_part04.txt, %40, 8 CPU/16 GB/3 h, partition ral only, script hpc/batch_cpu/submit_sersic_waveband_missing (skips complete tiles)388371— DR1 missing-SED sweep dr1_v2_new (euclid_dr1_sed_miss_v2new): 303 sep28 tiles with completed vis_lp 384040 — first SED on non-DECam ground bands (HSC/MegaCam/PS1), manifest sed_missing_20261003_v2new.txt, %10, 8 CPU/16 GB/3 h, partition ral only, script hpc/batch_cpu/submit_sersic_waveband_missing (skips complete tiles)388372— DR1 missing-SED retry probe (euclid_dr1_sed_miss_retry): tasks 0-49 of 1,252 earlier SED TIMEOUT/FAILED/partial tiles, manifest sed_missing_20261003_retry.txt, %10; send 50-1251 only if the probe does not mostly hang, 8 CPU/16 GB/3 h, partition ral only, script hpc/batch_cpu/submit_sersic_waveband_missing (skips complete tiles)398270— DR1 dr1_sep1 ex-top1000 vis_pix chained on vis_lp 397467 (euclid_dr1_vis_pix_ex_top1000_p01): immediate array 398270 over 256 lenses whose vis_lp had completed (%60) + 741 single-task jobs 398271-399020 each afterok:397467_N --kill-on-invalid-dep=yes; manifest vis_lp_ex_top1000_tonight_20261007.txt, Numba sparse CPU 8 CPU/16 GB/24 h, partition ral only, --exclude=euclid-ral-compute-[1-4] (left for another user's 252-CPU whole-node jobs) and --nice=500 so it never outranks pending work; driver submit_vis_pix_rest_chain.sh --batch=ex_top1000 (idempotent, re-run to catch stragglers)400282— DR1 dr1_sep1 SED fill 3 pass 1 part 01/4 (euclid_dr1_sed_fill3_p01): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move hung band aside, restart; band given up after 2 hangs via SED_SKIP_BANDS), manifest sed_fill_20261009/part01.txt (3,863 tiles over 4 parts: incomplete partial tiles fewest-missing-first, then none, then ex-top1000 with vis_lp; tiles still running in 397469/70/71/74 excluded), %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100400308— DR1 dr1_sep1 SED fill 3 pass 1 part 02/4 (euclid_dr1_sed_fill3_p02): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move hung band aside, restart; band given up after 2 hangs via SED_SKIP_BANDS), manifest sed_fill_20261009/part02.txt (3,863 tiles over 4 parts: incomplete partial tiles fewest-missing-first, then none, then ex-top1000 with vis_lp; tiles still running in 397469/70/71/74 excluded), %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100400309— DR1 dr1_sep1 SED fill 3 pass 1 part 03/4 (euclid_dr1_sed_fill3_p03): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move hung band aside, restart; band given up after 2 hangs via SED_SKIP_BANDS), manifest sed_fill_20261009/part03.txt (3,863 tiles over 4 parts: incomplete partial tiles fewest-missing-first, then none, then ex-top1000 with vis_lp; tiles still running in 397469/70/71/74 excluded), %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100400310— DR1 dr1_sep1 SED fill 3 pass 1 part 04/4 (euclid_dr1_sed_fill3_p04): submit_sersic_waveband_sed_fill_20261009 = 10-05 fill + 20-min stall watchdog (kill, move hung band aside, restart; band given up after 2 hangs via SED_SKIP_BANDS), manifest sed_fill_20261009/part04.txt (3,863 tiles over 4 parts: incomplete partial tiles fewest-missing-first, then none, then ex-top1000 with vis_lp; tiles still running in 397469/70/71/74 excluded), %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100402719— DR1 dr1_sep1 SED fill 4 pass 1 part 01/4 (euclid_dr1_sed_fill4_p01): indices 697-999 of sed_fill_20261009/part01.txt, replacing 400282's cancelled pending tasks; submit_sersic_waveband_sed_fill_20261010 = 10-09 fill + VIS given up after two hangs (exit 3, logged to sed_fill_20261009/given_up_bands.txt) and tiles already given up for vis skipped; %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100402720— DR1 dr1_sep1 SED fill 4 pass 1 part 02/4 (euclid_dr1_sed_fill4_p02): indices 527-999 of sed_fill_20261009/part02.txt, replacing 400308's cancelled pending tasks; submit_sersic_waveband_sed_fill_20261010 = 10-09 fill + VIS given up after two hangs (exit 3, logged to sed_fill_20261009/given_up_bands.txt) and tiles already given up for vis skipped; %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100402721— DR1 dr1_sep1 SED fill 4 pass 1 part 03/4 (euclid_dr1_sed_fill4_p03): indices 363-999 of sed_fill_20261009/part03.txt, replacing 400309's cancelled pending tasks; submit_sersic_waveband_sed_fill_20261010 = 10-09 fill + VIS given up after two hangs (exit 3, logged to sed_fill_20261009/given_up_bands.txt) and tiles already given up for vis skipped; %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100402722— DR1 dr1_sep1 SED fill 4 pass 1 part 04/4 (euclid_dr1_sed_fill4_p04): indices 803-862 of sed_fill_20261009/part04.txt, replacing 400310's cancelled pending tasks; submit_sersic_waveband_sed_fill_20261010 = 10-09 fill + VIS given up after two hangs (exit 3, logged to sed_fill_20261009/given_up_bands.txt) and tiles already given up for vis skipped; %25, 8 CPU/16 GB/6 h, partition ral only, --nice=100402909— DR1 SED fill 4 pass 2 (retry) part 01/4 (euclid_dr1_sed_fill4_retry_p01): indices 0-999 of sed_fill_20261009/part01.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100402910— DR1 SED fill 4 pass 2 (retry) part 02/4 (euclid_dr1_sed_fill4_retry_p02): indices 0-999 of sed_fill_20261009/part02.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100402911— DR1 SED fill 4 pass 2 (retry) part 03/4 (euclid_dr1_sed_fill4_retry_p03): indices 0-999 of sed_fill_20261009/part03.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100402912— DR1 SED fill 4 pass 2 (retry) part 04/4 (euclid_dr1_sed_fill4_retry_p04): indices 0-862 of sed_fill_20261009/part04.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100
Recent log — 5 of 136 entries
- 2026-10-10 — run — 402912 submitted: DR1 SED fill 4 pass 2 (retry) part 04/4 (euclid_dr1_sed_fill4_retry_p04): indices 0-862 of sed_fill_20261009/part04.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100
- 2026-10-10 — run — 402911 submitted: DR1 SED fill 4 pass 2 (retry) part 03/4 (euclid_dr1_sed_fill4_retry_p03): indices 0-999 of sed_fill_20261009/part03.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100
- 2026-10-10 — run — 402910 submitted: DR1 SED fill 4 pass 2 (retry) part 02/4 (euclid_dr1_sed_fill4_retry_p02): indices 0-999 of sed_fill_20261009/part02.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100
- 2026-10-10 — run — 402909 submitted: DR1 SED fill 4 pass 2 (retry) part 01/4 (euclid_dr1_sed_fill4_retry_p01): indices 0-999 of sed_fill_20261009/part01.txt; same script submit_sersic_waveband_sed_fill_20261010 and manifest; afterany on pass 1 400282/400308/400309/400310 (old, running tasks) and 402719-402722; complete tiles exit at once and tiles with vis given up are skipped, so it reruns only what is still incomplete; %25, 6 h, partition ral only, --nice=100
- 2026-10-10 — run — 400314 failed: DR1 SED fill 3 pass 2 (retry) part 04/4 (euclid_dr1_sed_fill3_retry_p04): same script and manifest part04.txt, afterany on pass 1 400282/400308/400309/400310 and the old fill 397469/70/71/74; complete tiles exit at once, so it reruns only what is still incomplete; %25, 6 h, partition ral only
active · ral · no issue yet
analytic_gaussian — Graphical and EP against a closed-form Gaussian posterior, at ensemble scale · no runs recorded · updated 2026-10-23
Registered: not recorded
Local: /mnt/c/Users/Jammy/Science/active/analytic_gaussian
RAL: /mnt/ral/jnightin/active/analytic_gaussian
Current position
Wave 1 (342413, 200 seeds at N=5) is accepted as the baseline: autofit EP is exact on the Gaussian leg, leg B sigma misses as pre-registered (78/200), collapse rate 0/200. On criterion 2, astra and an independent Opus review both favour an under-calibrated mu threshold (calibrated on five unrepresentative seeds) and found no defect in the minimal-EP control; both flag the per-site sigma>0 clip (analytic_ep_minimal.py:333) as the one audit target, and astra will not clear the control without an independent reconstruction of the EP fixed point. Next: run the planned diagnostic (archive/tasks/analytic_gaussian/minimal_ep_legb_mu_threshold.md) on a passing, an edge and the worst mu seed; the N=25 rung stays written but unsubmitted until it lands.
📋
Use the cortex skill. — resume analytic_gaussian: read PyAutoCortex projects/analytic_gaussian.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/analytic_gaussian/wiki/project/state.md and the assistant autofit_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
Recent log — 5 of 9 entries
- 2026-10-23 — note — second wave-1 finding closed (PyAutoFit#1713, PR #1714 merged, autofit_workspace_test#116): the
invalid value encountered in addwarnings on seeds 25 and 172 are a readout-only defect — under the mode projection an accepted truncated-normal scatter message sits ~19 sigma into its tail, its log_partition underflows to −inf (true log Z ≈ −187.9), the sigma evidence goes +inf and the EP log_evidence NaN from sweep 4; posteriors bit-identical to the ens_n5 sidecars. The library now names the term once per fit and returns an explicit NaN. The cure (tail-stable log_partition via log_ndtr) is message maths, filed in the Mind as draft/bug/autofit/truncated_normal_log_partition_tail_underflow.md for the human. - 2026-10-23 — note — criterion 6 amended (PyAutoFit#1711, PR #1712): the 19/400 zero-SUCCESS seed-legs of wave 1 split into 13 NO_CHANGE-only prior factors (zero-iteration Laplace convergence at the start point, projection accepted, closed-form match a 5e-5 / b 2e-5 on seed 140 leg A — a converged update, not stale; library reporting aligned, no EP maths changed) and 6 BAD_PROJECTION-only hierarchical factors (the #1405 scatter state, cured by projection="moments" #1656; seeds 29 and 71 re-run under moments give HierarchicalFactor SUCCESS 28 and 23). Criterion 6 now counts SUCCESS or NO_CHANGE as a completed update; the matching ensemble_seed.py any_factor_never_succeeded change is the human's to apply in the science checkout. Side finding: seed 29 leg B under moments still misses the sigma cell (a 0.225 / b 0.416) while sigma stays inside [q05, q95].
- 2026-10-14 — note — PyAutoFit#1656 (merged 2026-09-30, released 2026.10.7.1) ports the minimal-EP referee's moment matching as LaplaceOptimiser(projection="moments"); on analytic_gaussian.py leg B autofit EP sigma 6.3401 ± 2.4609 vs closed form 6.5667 ± 2.8832 (mode path 9.3722, BAD_PROJECTION=19). A wave-2 ensemble on the moments path needs a private PyAutoFit checkout on RAL: the shared mirror is 404b3e5f7 and lacks it. The criterion-2 mu-threshold diagnostic is still unrun.
- 2026-10-14 — note — output/ vaulted to z_vault/ on laptop and RAL (mv, nothing deleted), anticipating reruns; output/ left empty, results/ untouched; z_vault/ gitignored locally
- 2026-09-30 — note — astra (Codex gpt-6-astra) on criterion 2, verbatim: 'Neither is established as "wrong" by this result alone. I favour inadequate calibration of the threshold, with moderate confidence, but would not yet certify the control.' Decisive diagnostic: independently reconstruct the same EP factorisation at the fixed point on a passing, an edge and the worst mu seed and compare tilted vs q moments; audit the per-site sigma>0 clip (analytic_ep_minimal.py:333). Opus reviewer concurs independently (a_mu = 0.041*(E_ref[mu]-m0)/std_ref over 200 seeds, corr 0.98; calibration seeds 0-4 had |a_mu| 0.002-0.020 vs 95th pct 0.077). Full answer: PyAutoCortex#50 comment #50 (comment)
active · ral · PyAutoCortex#34
ep_toy_gaussian — NUTS versus three EP fits on the #1405 collapse toy at N=50 · no runs recorded · updated 2026-10-14
Registered: not recorded
Local: /mnt/c/Users/Jammy/Science/active/ep_toy_gaussian
RAL: /mnt/ral/jnightin/active/ep_toy_gaussian
Current position
Wave 1 (n50_seed42: 342639 + 342640_[0-2]) failed on infrastructure — EP x3 on EMFILE (fixed by PyAutoFit#1632/#1634, both in the RAL mirror at 404b3e5f7), NUTS on OOM at the 8 GB cap while XLA compiled the 50-factor window-adaptation scan, which grows with N (≥5.4 GB at N=20 locally). Wave 2 is prepared but not submitted: sample n50_seed42_w2 (free locally and on RAL; still needs its toy.SAMPLES entry), submit_nuts raised to 64 GB, submit_ep unchanged at 8 GB (peak ~590 MB). The collapse question at N=50 is unanswered; only n5_smoke speaks (EP RECOVER 3/3, NUTS 49.89/11.14). Next: human go, add the sample entry, push and submit NUTS then EP with SAMPLE=n50_seed42_w2 exported.
📋
Use the cortex skill. — resume ep_toy_gaussian: read PyAutoCortex projects/ep_toy_gaussian.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/ep_toy_gaussian/wiki/project/state.md and the assistant autofit_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
Recent log — 5 of 10 entries
- 2026-10-14 — note — Wave 2 still unsubmitted. RAL shared mirror 404b3e5f7 already carries the EMFILE fixes (#1632/#1634) wave 2 needs; it lacks #1655 (ProjectionException recovery), so a private PyAutoFit checkout is optional here but advisable.
- 2026-10-14 — note — output/ vaulted to z_vault/ on laptop and RAL (mv, nothing deleted), anticipating reruns; output/ left empty, results/ untouched; z_vault/ gitignored locally
- 2026-09-30 — note — RAL mirror PyAutoFit now 404b3e5f7 (2026-09-27), contains #1632 (b82fb3f69) and #1634 (fa2d540ac); 5 commits behind origin/main, none in blackjax/graphical/dynesty; deps at floor except anesthetic 2.8.14 < 2.9.0 (plots only). Sample n50_seed42_w2 is free locally and on RAL but must be added to toy.SAMPLES (scripts/toy.py) before submission, and SAMPLE must be exported to the job (hpc/sync submit does not forward laptop env)
- 2026-09-30 — note — wave-2 prep: 342639 NUTS OOM was XLA compilation of the vmapped window-adaptation scan, not sampling (sacct MaxRSS 8,180,536K at ReqMem 8G; py-spy on a local N=50 run sits in backend_compile_and_load; local probes N=10 3.2 GB peak / 132 s warmup compile, N=20 >=5.4 GB / 247 s, N=50 still compiling at 5 min) — submit_nuts --mem raised 8gb -> 64gb locally, uncommitted (est. peak 15-25 GB on RAL; ral nodes ~921 GB); EP peak ~590 MB, submit_ep stays 8gb
- 2026-09-24 — note — per-search overhead profile on this toy (2026-09-24, laptop, N=5, max_steps=3): 2.45 s per Dynesty factor search, ~1.47 s dynesty run_nested + ~0.98 s autofit wrapper (plots, a redundant second run_nested pass, samples writes); inside run_nested instance_from_vector is 57 % of the likelihood callback for a 1-parameter model. Speed-up prompt filed in the Mind: draft/refactor/autofit/ep_factor_search_wrapper_overhead.md
active · ral · ep_toy_gaussian#1
slope_hierarchy_scale — Hierarchical slope recovery with graphical and EP at N=25 to 50 · no runs recorded · updated 2026-10-14
Registered: not recorded
Local: /mnt/c/Users/Jammy/Science/active/slope_hierarchy_scale
RAL: /mnt/ral/jnightin/active/slope_hierarchy_scale
Current position
The N=25 graphical baselines are in: per-lens fits (342348) give a naive parent of 1.999 / 0.089 (deconvolved 0.085), and the joint NUTS fit (342350_0) recovers mean 1.994 [1.976, 2.014] and sigma 0.087 [0.074, 0.104] against truth 2.0 / 0.1; its summary JSON still needs writing offline from samples.csv (now under z_vault/sample_n25_seed42). The EP arm returned the prior on the Laplace projection (343299: hierarchical factor 0/50 SUCCESS). The cure has landed: PyAutoFit#1656 (merged 2026-09-30, released 2026.10.7.1) adds af.LaplaceOptimiser(projection="moments"), which takes the tight toy from 0 to 3/3 SUCCESS. The shared RAL PyAutoFit mirror is still 404b3e5f7 (2026-09-27, kept frozen for euclid_dr1) and lacks both #1655 and #1656, so any rerun needs a private PyAutoFit checkout on RAL. All earlier output is vaulted in z_vault/ on laptop and RAL, output/ is empty for the rerun. Next: rerun the N=25 EP arm with projection="moments" on a private PyAutoFit checkout and compare its scatter against the NUTS baseline.
📋
Use the cortex skill. — resume slope_hierarchy_scale: read PyAutoCortex projects/slope_hierarchy_scale.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/slope_hierarchy_scale/wiki/project/state.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
Recent log — 5 of 14 entries
- 2026-10-14 — note — output/ vaulted to z_vault/ on laptop and RAL (mv, nothing deleted), anticipating reruns; output/ left empty, results/ untouched; z_vault/ gitignored locally
- 2026-09-30 — note — graphical baselines recorded late at the 2026-09-30 check-in. 342348_[0-24] per-lens (one_by_one Nautilus, A100, 25/25, 6.9-8.8 min/task, 0 float32 truncations): naive parent mean 1.9992 +/- 0.0178 (SEM), std of medians 0.0889, deconvolved scatter 0.0852 (rms 1-sigma half-width 0.0255); truth inside the 1-sigma interval for 17/25 lenses; per-lens JSONs on RAL at results/one_by_one_sample_n25_seed42/dataset_*.json (not pulled; aggregate never written by the array). 342350_0 joint hierarchical NUTS: science completed despite the post-processing OOM — 500 warm-up (8h08m) + 1000 samples (7h17m), 1 chain, 0 divergences, acceptance 0.84, rhat_max 1.012, ess_min 133/302 params; parent mean 1.9936 [1.9761, 2.0136] 68%, sigma 0.0871 [0.0744, 0.1041] 68% (95% [0.0636, 0.1190]) vs truth 2.0/0.1 and realised draws 1.9917/0.0864 — the N=5 high-scatter bias (0.143) is gone at N=25. Computed from RAL samples.csv (graphical_nuts/f16c32e17f570603a1dc061f95792798/files); results/graphical_nuts_sample_n25_seed42.json never written (killed at 64 GB in latent/arviz post-processing) — pull samples.csv and write the summary offline
- 2026-09-30 — run — 342351_0 failed — wall 27:41: n25_scale_up: EP arm on the CPU partition, hpc/batch_cpu/submit_ep, max_steps 12; RAL PyAutoFit mirror 68ff9bd57 predates PyAutoFit#1580, so this arm runs without the stale-mask fixed-point fix — CANCELLED 2026-09-09 20:56:19 BST (sacct 1-03:41:06): log stops at the first Nautilus status line 2026-09-08 17:15:36, no factor finished; superseded by 342410 then 343299, output parked as ep_dead_342351; recorded late at the 2026-09-30 check-in
- 2026-09-30 — run — 342350_0 failed — wall 16:29: n25_scale_up: joint hierarchical NUTS fit, hpc/batch_gpu/submit_graphical — OUT_OF_MEMORY 2026-09-09 13:06:31 BST (sacct 0:125, cgroup oom-kill): NUTS warm-up 500 + sampling 1000 steps completed and samples.csv was written 12:14 (output/sample_n25_seed42/graphical/graphical_nuts/f16c32e17f570603a1dc061f95792798/files/samples.csv), then killed during latent samples / arviz post-processing; recorded late at the 2026-09-30 check-in
- 2026-09-30 — run — 342348_[0-24] finished — wall 3:22: n25_scale_up: one lens per array task, hpc/batch_gpu/submit_one_by_one, sample_n25_seed42 — all 25 tasks COMPLETED 2026-09-08 17:14-20:37 BST (sacct, 6:54-8:50 per task, serial on the GPU): per-lens one_by_one slopes are the baseline, e.g. dataset_0 2.0417 [2.0130, 2.0730] (truth 2.0305), dataset_24 2.0333 [2.0205, 2.0463] (truth 2.0439); recorded late at the 2026-09-30 check-in
active · both · slope_hierarchy_scale#2
ic50_workspace — EP against the graphical joint fit on IC50 dose-response data, scaling up · no runs recorded · updated 2026-10-14
Registered: not recorded
Local: /mnt/c/Users/Jammy/Science/active/ic50_workspace
RAL: /mnt/ral/jnightin/active/ic50_workspace
Current position
The scale ladder is in. EP passes N=5/10/25 with coef_mean within 3σ and cost growing about linearly per sweep (26/50/140 s); it died at N=50 on a projection assert. That fix has landed: PyAutoFit#1655 (merged 2026-09-30, released 2026.10.7.1) raises ProjectionException, which EP factor_step now recovers from instead of killing the run. Graphical is cheaper but overconfident beyond N=25, and EP hill_coef widths are not yet comparable because util.py reports factor-message widths. The shared RAL PyAutoFit mirror is still 404b3e5f7 (2026-09-27, kept frozen for euclid_dr1) and lacks both #1655 and #1656, so any rerun needs a private PyAutoFit checkout on RAL. All earlier output is vaulted in z_vault/ on laptop and RAL, output/ is empty for the rerun. Next: rerun the EP ladder on a private PyAutoFit checkout with output on and a fixed seed, report hill_coef from the mean field, then give ep_lbfgs_jax a witness as the scale lever.
📋
Use the cortex skill. — resume ic50_workspace: read PyAutoCortex projects/ic50_workspace.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/ic50_workspace/NOTES.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
Recent log — 5 of 12 entries
- 2026-10-14 — note — output/ vaulted to z_vault/ on laptop and RAL (mv, nothing deleted), anticipating reruns; output/ left empty, results/ untouched; z_vault/ gitignored locally
- 2026-09-30 — note — Ladder read: EP s/sweep 26/50/140/~300 s at N=5/10/25/50 vs graphical wall 21/42/119/281 s. EP hill_coef widths flat at ~0.8 because util.py reports factor-message (likelihood) widths and the global factor freezes hill_coef; graphical widths shrink but fall to 56/75 and 65/150 within 3 sigma at N=25/50 (overconfident). coef_mean passes 3 sigma for both methods on every completed rung.
- 2026-09-30 — note — 342411 N=50 failure diagnosed: AssertionError at autofit/messages/abstract.py:316 while projecting the global factor after EP sweep 4 (mirror 66f9f8d5d); the 4th global search had max logL -337 (vs -69 at sweep 3) and logz +/- nan. Library defect: project asserts on non-finite suff_stats and factor_step (optimiser.py:150) does not catch AssertionError, so one bad projection kills the run; no later commit touches the path. A local rerun at 66f9f8d5d converged in sweep 3 without failing (unseeded Dynesty, non-deterministic). Mind bug prompt draft/bug/autofit/ep_project_nonfinite_suff_stats_ic50_n50.md
- 2026-09-09 — run — 342411 failed — wall 0:38: ep_scale_up: EP scale ladder, sim rungs 5/10/25/50, nlive 150 max_steps 12 — finished 2026-09-09 23:19:29 BST (sacct COMPLETED 00:38:49, the ladder catches rung failures): rungs N=5/10/25 passed the global coef_mean assertions; rung N=50 FAILED — AssertionError: assert np.isfinite(suff_stats).all() in autofit/messages/abstract.py:316 project (via mapper/prior/abstract.py:237), error.342411.err
- 2026-09-09 — run — 342412 finished — wall 0:07: ep_scale_up: graphical scale ladder, sim rungs 5/10/25/50, nlive 150 — COMPLETED 2026-09-09 22:48:42 BST (sacct 00:07:59): all four rungs N=5/10/25/50 passed the global coef_mean assertions (within 3σ); results/graphical_ladder_n*_summary.json
active · ral · PyAutoCortex#36
autolens_inference — Do the PyAutoLens inference backends agree on the HST SLaM posterior · no runs recorded · updated 2026-10-21
Registered: not recorded
Local: /home/jammy/Code/PyAutoLabs/lens/autolens_inference
RAL: /mnt/ral/jnightin/inference/autolens_inference
Current position
Clean slate on pin 2026.10.17.1, campaign hst-imaging-mass-samplers, state read 2026-10-19 15:49:05Z. Nothing is queued: on the human's decision, 412541 (jax_cpu sparse) and 412565 (numba_cpu sparse) were cancelled ~15:51Z with partial stages left on disk (no CPU fixed-light row exists). 412535 and 412537 (A100) and 412539 (jax_cpu dense) completed; numba_cpu dense dropped. The project is on the flat layout (autolens_inference#36, c3ca4de; output/autolens/..., rows at results/slam/imaging/hst/<variant>/<config>/), and the wave-1 attempt-1 material is set aside. Next, each needing the human's go: confirm keeping prepared/; a Mind bug task and laptop repro for the jax_gpu sparse fixed-light mass_total[1] divergence; rerun the cancelled sparse remainder (jax_cpu sparse seed 1 from source_pix[2], the lightfixed mass_total[1] legs); GPU wave 1 attempt 2 (nautilus+emcee); commit the clean-slate rows in a project PR. No scientific acceptance implied.
📋
Use the cortex skill. — resume autolens_inference: read PyAutoCortex projects/autolens_inference.md (Now, Runs, Log) and then /home/jammy/Code/PyAutoLabs/lens/autolens_inference/wiki/project/state.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
Recent log — 5 of 46 entries
- 2026-10-21 — note — autolens_inference#43 (likelihood-lifecycle phase 0C, merged 2026-10-21 as 07880c2) moved the rows: results/slam/imaging/hst/<variant>/<config>/stages_seed<n>.* are now results/imaging/<rectangular|delaunay>/hst/validation/<default|light_fixed>/<setup_id>/seed_<n>/stages.*; results/searches/point_source/<sampler>/... are results/point_source_source/source_plane/simple/samplers/default/<sampler>/explicit/<setup_id>/seed_<n>/; results/archive/<dated> is archive/YYYY/MM/<dated>/. Old → new map: catalogue/migration_2026_10.json. Entries above this line cite the old paths. #44 (bdc8b2d) exports dashboard/walltime_validation.json (walltime-validation@1, 56 records) to PyAutoPulse.
- 2026-10-19 — note — 412539 (jax_cpu dense) elapsed from sacct, read 2026-10-19 ~17:25Z by the session architect: 412539_0 1-05:59:53 (ended 2026-10-18T23:46:01), 412539_1 1-02:57:03 (ended 2026-10-18T20:43:11). This supplements the 2026-10-19 finished entry, which carries no wall. Its rows have still not been read.
- 2026-10-19 — run — 412543 failed — wall 0:47: hst-imaging-mass-samplers wave 1 (cleanslate_2026_10_17_wave1, submit_hst_cleanslate_mass_samplers_jax_gpu_dense): 24-arm slice nautilus/emcee/nuts/smc x sampler seeds 0-2 x both light modes on the base-seed-0 prepared problems of hpc_a100_jax_gpu_dense_fp64, array 0-23%4, 6 h containment, --dependency afterok:412535_0, pin 2026.10.17.1; human-authorised 2026-10-17 — sacct read 2026-10-19 ~17:25Z by the session architect: tasks _0.._17 all FAILED on 2026-10-17 (ends 18:51:38-20:32:26Z, elapsed 00:04:15-00:47:10; wall given is the longest task); _18 CANCELLED after 00:04:00; _[19-23] CANCELLED unstarted 2026-10-17T20:01:03Z. Failure classes as in the 2026-10-17 note (NUTS/SMC A100 OOM; nautilus/emcee row-writer TypeError). Its output and the 8 failed NUTS/SMC rows were moved aside on 2026-10-19 to aside/autolens_inference_cleanslate_wave1_attempt1/ (RAL and laptop)
- 2026-10-19 — note — HUMAN DECISION 2026-10-19, verbatim: "I want us to check the on going inference campaign on RAL, but I am ready for us to update folder structures on RAL and locally and clean things up so cancel on going jobs and we'll work on the clean up". Scheduler read 15:49:05Z: only 412541 and 412565 running; both scancelled ~15:51Z, queue empty; 412539 had COMPLETED. On "yes, go", autolens_inference#36 (flatten, closes #35) merged 15:53:33Z as c3ca4de; the RAL checkout fast-forwarded b91ce9c -> c3ca4de and the laptop to c3ca4de, after a full hpc/sync pull. flatten_cleanslate.py --apply: RAL 28 moves, 55 rewrites; laptop mirror 28 moves, 38 rewrites; both EXIT 0. New layout: output/autolens/imaging/hst/slam/<variant>/<config>/seed_<n>/<stage>/, output/autolens/prepared/<config>/seed_<n>/<variant>/, rows at results/slam/imaging/hst/<variant>/<config>/ (no __cleanslate_2026_10_17 suffix). RAL and laptop each hold 1974 files with identical paths and sizes; hpc/sync status dry run transfers 0. Human, verbatim: "check some older chats, think we dont need prepared, experiments and experiment_sources to exist anymore? Can we archive them somewhere separate to makr things easier to follow. Then do the apply". The wave-1 attempt-1 material (output/experiments 341M, output/experiment_sources 76M, results/searches/experiments with 8 failed NUTS/SMC rows) moved to /mnt/ral/jnightin/inference/aside/autolens_inference_cleanslate_wave1_attempt1/ on RAL and /mnt/c/Users/Jammy/Science/inference/aside/autolens_inference_cleanslate_wave1_attempt1/ on the laptop. prepared/ KEPT as the agent's proposal, not yet confirmed by the human: wave.py needs its manifests, and so does the sparse fixed-light repro. Partial stage dirs of the cancelled legs remain; they are not failures. Execution completion only. No compute submitted.
- 2026-10-19 — run — 412565 failed — wall 44:52: hst-imaging-mass-samplers clean slate: numba_cpu sparse SLaM base (64gb, 8 CPUs), --lens-light both, seeds 0-1, from autolens_inference main b91ce9c (PR #32), pin 2026.10.17.1, output root output/cleanslate_2026_10_17, results tag cleanslate_2026_10_17, prepared export; --stages-from hpc_a100_jax_cpu_dense_fp64:source_lp shares source_lp[1] from the jax_cpu dense chain with an in-job wait (hpc/wait_for_stage.sh); rows mark that stage shared and exclude its wall from the numba chain wall; submitted 2026-10-17T18:58Z by the architect; human decisions 2026-10-17 (8 CPUs, shared stage, submit now) — CANCELLED by scancel ~15:51Z on the human's decision, not a run failure (sacct CANCELLED, elapsed 1-20:52:11). Complete at cancel (.completed markers): seeds 0 and 1 slam_base light[1], source_pix[1], source_pix[2], mass_total[1] (source_lp[1] shared from jax_cpu dense via --stages-from). Incomplete: slam_base_lightfixed mass_total[1] on both seeds (running). Partial stage dirs left on disk under output/autolens/imaging/hst/slam/<variant>/hpc_a100_numba_cpu_sparse_fp64/seed_<n>/ (flat layout, autolens_inference#36)
active · both · autolens_inference#4
autofit_inference — PyAutoFit search benchmark on gaussian_x3: which searches find the right answer fastest · no runs recorded · updated 2026-10-16
Registered: 2026-10-08
Local: /home/jammy/Code/PyAutoLabs/fit/autofit_inference
RAL: /mnt/ral/jnightin/inference/autofit_inference
Current position
B5c-2a (2026-10-16, autofit_inference#17): the correlated_nd pilot (RAL 410866) is pulled from the private copy and judged under correlated_nd@1 (1040 expected = 749 completed rows + 177 failed + 4 oom + 0 missing + 110 deferred; 307 accepted, 623 rejected; ranks nothing, nothing ranked across families), and correlated_nd@2 is frozen in autofit_inference 44c5e77f1fb0 (protocol section 11 A2) before any scored row: sigma-ratio bands per task (evidence [0.85, 1.1] on 79 rows, posterior kept at [0.5, 2.0] on 1 row), ESS/R-hat bars kept, 75 timed + 8 probe-first cells, 21 cells deferred with reasons (incl. d50 Nautilus n_live_100 by the cap rule and all d100 Nautilus by A2), memory 2x the pilot peak (d50 Nautilus n_live_400 16 G, d100 cold NUTS 17 G). Calibration: autofit_inference results/calibration/correlated_nd_wave1.json. gaussian_x3 wave 2 stays closed under gaussian_x3@2. Next: B5c-2b (draft/research/autofit_inference/search_ext_b5c2b_scored_wave_submit.md) adds realisations 2-3 and their exact references, the 12 450-run scored-wave manifest and its probe, and submits on --partition=ral from a private pinned copy; nothing is submitted until then.
📋
Use the cortex skill. — resume autofit_inference: read PyAutoCortex projects/autofit_inference.md (Now, Runs, Log) and then /home/jammy/Code/PyAutoLabs/fit/autofit_inference/wiki/project/state.md and the assistant autofit_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
Recent log — 5 of 28 entries
- 2026-10-16 — result — correlated_nd pilot RAL 410866 judged under correlated_nd@1 (B5c-2a, autofit_inference#17): results/ pulled from the private copy inference/autofit_inference_b5c (930 rows; exact references byte-identical); 1040 expected = 749 completed rows + 177 failed + 4 oom + 0 missing + 110 deferred, reconciled with sacct; 307 accepted, 623 rejected; ranks nothing. Failures: 80 Emcee/Zeus (d20, banana) crashed in the harness's chain diagnostics after the search (relabelling by g0.centre; fixed, no verdict moves); 60 d100 Nautilus (n_live_100: needs n_live > D; n_live_200/400: nautilus HDF5 checkpoint cannot store a D = 100 bound); 10 d100 warm NUTS (provider); 27 at the 4 h cap (d50 Nautilus n_live_100, d50 warm NUTS). The 4 OUT_OF_MEMORY tasks 263/273/304/383 (MaxRSS 7.6/7.7/3.1/8.4 G at 8 G) had been given cap stubs by the exit-137 path and are now an oom attempt class. A2 ruling: the OOM cells d50 Nautilus n_live_100 and d50 warm NUTS are deferred by the cap rule (no attempt ended on its own); d50 Nautilus n_live_400 gets 16 G and d100 cold NUTS 17 G (2x the measured peak, 8 G for an OOM class); d100 Nautilus n_live_200 (and 100, 400) is deferred (fails at the pin before sampling, affordability unmeasured); an oom run is rerun once at 2x. correlated_nd@2 frozen in autofit_inference 44c5e77f1fb0 (44c5e77f1fb0779f675a8f51247d812183e93b48) before any scored row; nothing submitted.
- 2026-10-16 — run — 410866_[0-929] finished — wall 4:16: correlated_nd@1 pilot (B5c), resubmission of 409957 from the PRIVATE copy /mnt/ral/jnightin/inference/autofit_inference_b5c (delivered files sha256-verified from 245a3ef, own copy of the pin): 930 runs, 10 seeds x realisation 1 x 93 planned cells, one run per task, 4 h process cap, %100; pin df6d218f10cd; pull results/ from the private copy in B5c-2 — 926/930 COMPLETED, 4 OUT_OF_MEMORY at 8 G, all corr_gaussian_d50: task 263 nautilus n_live_100 numpy seed 3, 273 nautilus n_live_100 jax_cpu seed 3, 304 nautilus n_live_400 numpy seed 4, 383 blackjax_nuts_warm warmup_200_samples_500 jax_cpu seed 3; not rerun in stage 1 (human decision), B5c-2 raises their memory or excludes them; results unpulled in the private copy, pulled and judged in B5c-2
- 2026-10-15 — run — 410866_[0-929] submitted: correlated_nd@1 pilot (B5c), resubmission of 409957 from the PRIVATE copy /mnt/ral/jnightin/inference/autofit_inference_b5c (delivered files sha256-verified from 245a3ef, own copy of the pin): 930 runs, 10 seeds x realisation 1 x 93 planned cells, one run per task, 4 h process cap, %100; pin df6d218f10cd; pull results/ from the private copy in B5c-2
- 2026-10-15 — run — 409957_[0-929] failed — wall 0:06: correlated_nd@1 pilot (B5c, second model family; protocol first committed 9ec37795f404, A1 before any row): 930 runs, 10 seeds x realisation 1 x 93 planned cells over corr_gaussian_d20/d50/d100 + rosenbrock_d20, one run per task, 4 h process cap, %100; submit_correlated_nd_pilot; pin df6d218f10cd; pulled in B5c-2 — every task failed at import (ModuleNotFoundError _autofit_inference_cli): the shared RAL clone had been switched to an old main by another session (stash pre-main-switch 19:24 UTC) and lost the CLI and config/ as the array started; 800 tasks exited, the rest scancelled; no row written; resubmitted from the private copy inference/autofit_inference_b5c as 410866
- 2026-10-15 — run — 409957_[0-929] submitted: correlated_nd@1 pilot (B5c, second model family; protocol first committed 9ec37795f404, A1 before any row): 930 runs, 10 seeds x realisation 1 x 93 planned cells over corr_gaussian_d20/d50/d100 + rosenbrock_d20, one run per task, 4 h process cap, %100; submit_correlated_nd_pilot; pin df6d218f10cd; pulled in B5c-2
active · ral · autofit_inference#8
euclid_detailed — Advanced SLaM remodel of 31 euclid_dr1 lenses with residuals · no runs recorded · updated 2026-10-15
Registered: 2026-10-15
Local: /mnt/c/Users/Jammy/Science/active/euclid_detailed
RAL: /mnt/ral/jnightin/active/euclid_detailed
Current position
Born 2026-10-15. Repo, 31 datasets, output_dr1 (euclid_dr1 fits) and scripts/slam.py ready; test mode passes both stages. Next: hpc/sync push, then submit_slam_chain.sh --sample=source_residuals and --sample=lens_light_residuals on RAL (partition ral).
📋
Use the cortex skill. — resume euclid_detailed: read PyAutoCortex projects/euclid_detailed.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/euclid_detailed/wiki/project/state.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
cowls — euclid_detailed SLaM pipeline on COWLS JWST strong lenses, from COSJ100024+015334 · no runs recorded · updated 2026-10-15
Registered: 2026-10-15
Local: /mnt/c/Users/Jammy/Science/active/cowls
RAL: /mnt/ral/jnightin/active/cowls
Current position
Born 2026-10-15. Repo PyAutoLabs/cowls (private), M25 COSJ100024+015334 dataset (4 bands) and scripts/slam.py ready; test mode passes both stages on F444W. Next, after the RAL PyAuto install is updated: hpc/sync push, then submit_slam_chain.sh --sample=M25 (F444W, partition ral).
📋
Use the cortex skill. — resume cowls: read PyAutoCortex projects/cowls.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/cowls/wiki/project/state.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
cowls_ansofi — autolens_inference Delaunay-1250 numba CPU SLaM on COSJ100024+015334, F444W only · no runs recorded · updated 2026-10-19
Registered: 2026-10-19
Local: /mnt/c/Users/Jammy/Science/active/cowls_ansofi
RAL: /mnt/ral/jnightin/active/cowls_ansofi
Current position
Born 2026-10-19. Repo PyAutoLabs/cowls_ansofi (private), M25 COSJ100024+015334 F444W dataset and scripts/slam.py ready; test mode passes all five searches (144 s, numba sparse on every stage). Next: hpc/sync push-submit cpu submit_slam (8 CPU, 64 GB, 5 days, partition ral).
📋
Use the cortex skill. — resume cowls_ansofi: read PyAutoCortex projects/cowls_ansofi.md (Now, Runs, Log) and then /mnt/c/Users/Jammy/Science/active/cowls_ansofi/wiki/project/state.md and the assistant autolens_assistant's AGENTS.md; tell me where I left off and what I said I would do next. Submit nothing and log nothing until I say.
Runs (0)
nothing on the cluster
euclid — The original Euclid DR1 grade-AB catalogue project, superseded by euclid_dr1 · dormant · no runs recorded · updated 2026-10-15
Registered: not recorded
Local: /mnt/c/Users/Jammy/Science/active/euclid
RAL: /mnt/ral/jnightin/active/euclid
Current position
Dormant: superseded by euclid_dr1 (euclid_dr1_prelim is defunct). The tree is kept under Science/active and RAL active/euclid (category layout, 2026-10-15); no question is open.
Runs (0)
nothing on the cluster
Recent log — 4 of 4 entries
- 2026-10-15 — note — superseded by euclid_dr1 (euclid_dr1_prelim defunct) per the human; stays dormant; folder under Science/active and RAL active/; ral_root repointed to the category layout (PyAutoCortex#67)
- 2026-08-28 — note — question: Do we recover Sersic indices or is prior pile-up artefact — planned, never run [archive/tasks/euclid/sersic_index_recovery.md]
- 2026-08-28 — note — question: Can a fitted Euclid lens be resimulated and recovered — planned, never run [archive/tasks/euclid/resimulate_fitted_lens_simulator.md]
- 2026-08-28 — note — question: How robust are magnification estimates under model match and mismatch — planned, never run [archive/tasks/euclid/magnification_robustness.md]
dormant · both · no issue yet
5 retired
- inference_programme — The first inference programme, retired and restarted as autolens_inference — retired 2026-09-07: restarting from scratch: the project setup was unsatisfactory and it informed how the Cortex manages projects; RAL outputs stashed at /mnt/ral/jnightin/inference_programme_retired_2026-09-07
- euclid_dr1_prelim — Refit the ten DR1-prelim tiles and rebuild the catalogue on RAL — retired 2026-09-25: retire euclid_dr1_prelim; tree moved to z_vault by the human
- slope_hierarchy — Hierarchical slope recovery at N=5, wrapped up and succeeded by slope_hierarchy_scale — retired 2026-09-08: wrapped up 2026-07-22 at N=5; succeeded by slope_hierarchy_scale; tree vaulted under z_projects_complete/
- euclid_sersics — Why does the lens-light Sersic index pile up at the n=5 prior edge — retired 2026-09-25: retire euclid_sersics as well; local tree deleted 2026-10-14 (no actionable results)
- pj011646_final — Final masked PJ0116-24 power-law and multipole subhalo comparison — retired 2026-10-14: complete; tree moved to z_projects_complete 2026-10-14
| Project | Status | Note |
|---|---|---|
| profiling | retired | retired 2026-10-14: human moved the tree to z_vault 2026-10-14 |
| cowls_diana | retired | retired 2026-10-14: human moved the tree to z_vault 2026-10-14 |
| subhalo_simulations | retired | retired 2026-10-14: superseded by subhalo_validation; tree moved to z_vault 2026-10-14 |
| pj011646 | retired | retired 2026-10-02: Human requested both PJ011646 and aris_PJ011646 archived intact to z_vault on laptop and RAL; superseded by pj011646_final |
| concr | retired | retired 2026-10-14: human moved the tree to z_vault 2026-10-14 |
Boards: Organism · Brain · Mind · Memory · Eyes · Ears · Heart · Hands · Pulse · Insight · Dna · Nerves · Broca · Gut