Archive 4 checkpoints (1-4)

paper-releases/pbisc-diffusion/updates/20260928T225702Z-paired-ft-scsemi-delta/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/giroux2022/fewshot_n5/draw0_scsemi5/README.md

@@ -0,0 +1,73 @@
+# score_ft.py output: giroux2022, 11 scored donors
+
+Arms: scsemi5.  Every arm is scored on the same donors against the same observed reference.
+
+## Inputs and fixed contracts
+- Observed reference: `/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/giroux2022/reference.npz` (sha dd0152f7bde4), built from datadir:/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/data/giroux2022;
+  observed cells sum exactly to the features pseudobulk of every donor. Observed labels: /home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/giroux2022/observed_labels/labels/labels.parquet.
+- Generated cells: `generated_counts` read out once with Poisson(seed 20260916) over the **whole NPZ in row order**
+  (`poisson_readout.poisson_integerise`), then the scored donors' rows. These integer counts are used for labels, embedding,
+  depth and the contrast. Pseudobulk uses the summed `generated_counts` (rates), as `score_generated.py`.
+- Labels: frozen f01common scANVI (`predict_scanvi.py`, CPU, zero query epochs), six-type gate ['T cell', 'B cell', 'Monocyte', 'NK cell', 'DC', 'Other']; query genes =
+  the 1,006-gene annotation list minus genes the dataset's sc assay did not measure in every sample (1006 genes).
+- Scoring genes: 6k panel genes measured by both assays in every sample of the dataset (5563 genes).
+
+## Definitions
+1. **Composition**: fraction of each donor's cells per six-type label (all generated cells; all observed QC cells).
+   `comp_l1` = sum over the six labels of |generated - observed| (immicro_common.l1). `comp_l1_obs_loo_null` = L1 of the
+   donor's observed composition to the mean of the *other scored donors'* observed compositions. `composition_top1` =
+   fraction of donors whose generated composition is closest (L1) to their own observed composition.
+2. **Pseudobulk**: `pb_r` = Pearson r of CPM-log1p(summed generated_counts) vs CPM-log1p(observed pseudobulk) on the scoring
+   genes (evaluate_realbulk.cpm_log1p / paired_r). `pb_top1` = the donor's generated pseudobulk is closest (Pearson) to its own
+   observed pseudobulk among the 11 scored donors (immicro_common.retrieval = evaluate_realbulk.top1 arithmetic).
+3. **Precision / coverage / density** (stage_b_embed.py functions): per-cell CP10K over the scoring genes + log1p; PCA-50
+   (randomized, random_state 20260925) and k = 10 real-real k-NN balls fitted on **R1 of the scored donors only**.
+   Reference rule (independent of fold and arm): per donor, cells in canonical order (library_id+barcode), one
+   permutation from `default_rng([20260925, uint64(sha256(donor)[:8])])`; R1 = first n, R2 = next n,
+   n = min(150, n_cells // 2) (1650 R1 cells here). Generated query: per donor n cells (the donor's R1 size),
+   without replacement, `default_rng([20260926, same key])`. Precision = fraction of generated cells inside any R1 ball
+   (Kynkaanniemi 2019); coverage = fraction of R1 cells whose ball holds a generated cell; density = Naeem 2020.
+   Per-donor precision = that donor's generated cells; per-donor coverage = that donor's R1 cells (the pooled values are
+   their n-weighted means; plain means when every donor has 150 reference cells). R2 (held-out real cells of the same donors) is the
+   perfect-generator control: precision 0.9988, coverage 1.0000.
+   **Depth matched**: stage_b_embed depth matching run pairwise (R1 vs this arm): per donor and within-donor rank of detected
+   genes, both cells are thinned by UMI sampling without replacement to the smaller detected-gene count; PCA kept from
+   un-thinned R1, radii recomputed on thinned R1; fresh rng(20260929) per arm. R2 control: precision
+   0.9982, coverage 1.0000.
+4. **Monocyte %** = 100 x Monocyte fraction (all cells) per donor; **median UMI per cell** = median over cells of the
+   Poisson-read (generated) / observed counts summed over the scoring genes (per donor, and pooled).
+5. **Group contrast** (T, B, Monocyte, NK): case groups ['severe'] (n = 6) vs control groups ['mild'] (n = 5); rule: explicit --case-groups / --control-groups; groups from features groups.
+   Genes per type = scoring genes detected (count > 0) in >= 1 % of the scored donors' observed
+   cells of that type. Primary statistic `mean_cp10k`: per-cell CP10K over the scoring genes; per donor the mean over its
+   cells of the type (donors with >= 1 such cell); group mean = mean over the group's donors (donor-balanced);
+   logFC = log2(case mean + 1) - log2(control mean + 1). Secondary `pseudobulk_log1p_cpm`: the SLE benchmark evaluator's
+   statistic (donor x type summed counts, log1p CPM over the scoring genes, mean case - mean control;
+   evaluate_generated_vs_truth.group_log_fold_change). Both are computed on the generated cells (their frozen labels) and on
+   the observed cells; `r_generated` = Pearson r of the two logFC vectors. Broadcast baseline `r_broadcast`: the same
+   observed vector against the logFC of the input bulk (condition NPZ `bulk_counts` of the arm, scoring genes, per-donor
+   CP10K resp. log1p CPM, group means). `partial_r_generated_given_broadcast`: r of the OLS residuals of observed and
+   generated logFC on the bulk logFC (residual_specificity.py).
+6. **Top-50 DEG overlap** (contrast.csv rows `statistic = welch_top50_deg`; summary `deg_top50`): the SLE benchmark's score (`total_top50_overlap` of
+   `bulk2sc-paper/scripts/evaluate_generated_vs_truth.py`, Welch backend: the "top-50 DEG 81 / 84" of the final model), run
+   with the evaluator's own functions on the same donor x type pseudobulks as the contrast. Per cell type, observed and
+   generated separately: donor x type summed counts on all 5563 scoring genes (donors of the two groups with
+   >= 1 cell of the type) -> `welch_bh_deg` (log1p CPM, two-sided Welch t-test case vs control, Benjamini-Hochberg over the
+   genes with a defined p-value; needs >= 2 donors per group) -> `top_deg_genes` (FDR < 0.05, descending |logFC|,
+   first 50; shorter when fewer genes pass). `top50_overlap` = |observed list & generated list|; `top50_jaccard`
+   the evaluator's Jaccard. Summary `sum` = total over the scored types (the quantity reported as 81), `max` = the sum of
+   the observed list lengths (the "of 262"), `sum_4type` = T + B + Monocyte + NK. A type is scored when its observed test
+   is defined; an arm whose generated test is undefined for a scored type gets overlap 0. Types: T cell, B cell, Monocyte, NK cell, DC
+   (aggregates: /home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/giroux2022/deg_reference.npz).
+   **Broadcast analogue** (`top50_overlap_broadcast`, as `sle_broadcast.py` `broadcast_input_bulk`; not part of the
+   published benchmark): the same test on the arm's condition bulk_counts (scoring genes, the 6 + 5
+   donors), one whole-PBMC top-50 list, compared with every type's observed list.
+
+## Files
+- `per_donor.csv`: one row per arm x donor (all per-donor quantities above).
+- `summary.json`: per arm the donor means / pooled values, provenance (NPZ sha, receipt check, generation metadata,
+  Poisson nnz of the whole NPZ, annotation contract), and the reference block (PCA, R2 control).
+- `contrast.csv`: per arm x cell type x statistic (`mean_cp10k`, `pseudobulk_log1p_cpm`: logFC r; `welch_top50_deg`:
+  top-50 DEG overlap, T / B / Monocyte / NK / DC; columns of the other kind are empty).
+- `arms/<arm>/labels/` (predict_scanvi.py labels + manifest), `arms/<arm>/generated_cells.parquet` (NPZ row, donor, label,
+  UMI on the scoring genes, P/C subsample flag), `arms/<arm>/deg_top50_lists.csv` (observed, generated and broadcast
+  top-50 gene lists, Ensembl ids, rank order).

paper-releases/pbisc-diffusion/updates/20260928T225702Z-paired-ft-scsemi-delta/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/giroux2022/fewshot_n5/draw0_scsemi5/contrast.csv

@@ -0,0 +1,14 @@
+arm,celltype,statistic,n_genes,r_generated,r_broadcast,partial_r_generated_given_broadcast,r_generated_vs_broadcast,n_case_donors_obs,n_control_donors_obs,n_case_donors_gen,n_control_donors_gen,n_cells_obs,n_cells_gen,deg_scored,top50_overlap,top50_jaccard,obs_top_list_size,gen_top_list_size,n_sig_obs,n_sig_gen,top50_overlap_broadcast,top50_jaccard_broadcast,broadcast_top_list_size,n_sig_broadcast
+scsemi5,T cell,mean_cp10k,2884,0.4722392025248643,0.5530162885771299,0.3355642876447728,0.38797538667457393,6,5,6,5,59044,9284,,,,,,,,,,,
+scsemi5,T cell,pseudobulk_log1p_cpm,2884,0.37226163006287905,0.7491008014749582,0.16703861462904504,0.359078250633904,6,5,6,5,59044,9284,,,,,,,,,,,
+scsemi5,B cell,mean_cp10k,3028,0.2618496573358809,0.46079426611859237,0.15939315076066354,0.27291413318063185,6,5,6,5,9139,749,,,,,,,,,,,
+scsemi5,B cell,pseudobulk_log1p_cpm,3028,0.16014760499406513,0.6336532411649639,0.03548517701741822,0.2103833069432783,6,5,6,5,9139,749,,,,,,,,,,,
+scsemi5,Monocyte,mean_cp10k,3050,0.40339497853778705,0.6395437766126884,0.2254605912761656,0.38008110129279193,6,5,6,5,19384,4192,,,,,,,,,,,
+scsemi5,Monocyte,pseudobulk_log1p_cpm,3050,0.22502197267148633,0.7472442358121685,0.05683593878381537,0.2522238689049979,6,5,6,5,19384,4192,,,,,,,,,,,
+scsemi5,NK cell,mean_cp10k,2935,0.3516321255863767,0.4632733189911659,0.21988451783327412,0.3678886167466043,6,5,6,5,11027,1815,,,,,,,,,,,
+scsemi5,NK cell,pseudobulk_log1p_cpm,2935,0.1993803268280498,0.6651655681008249,0.030987243561886684,0.2662152952390908,6,5,6,5,11027,1815,,,,,,,,,,,
+scsemi5,T cell,welch_top50_deg,5563,,,,,6,5,6,5,59044,9284,True,1,0.07692307692307693,13,1,13,1,5,0.35714285714285715,6,6
+scsemi5,B cell,welch_top50_deg,5563,,,,,6,5,6,5,9139,749,True,0,0.0,11,1,11,1,5,0.4166666666666667,6,6
+scsemi5,Monocyte,welch_top50_deg,5563,,,,,6,5,6,5,19384,4192,True,0,0.0,6,0,6,0,5,0.7142857142857143,6,6
+scsemi5,NK cell,welch_top50_deg,5563,,,,,6,5,6,5,11027,1815,True,0,0.0,14,0,14,0,5,0.3333333333333333,6,6
+scsemi5,DC,welch_top50_deg,5563,,,,,6,5,5,5,1558,75,True,0,0.0,7,0,7,0,5,0.625,6,6

paper-releases/pbisc-diffusion/updates/20260928T225702Z-paired-ft-scsemi-delta/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/giroux2022/fewshot_n5/draw0_scsemi5/per_donor.csv

@@ -0,0 +1,12 @@
+arm,donor,group,n_generated_cells,n_observed_cells,n_reference_cells,gen_T cell,obs_T cell,gen_B cell,obs_B cell,gen_Monocyte,obs_Monocyte,gen_NK cell,obs_NK cell,gen_DC,obs_DC,gen_Other,obs_Other,comp_l1,comp_l1_obs_loo_null,pb_r,pb_own_rank,pb_top1,precision,coverage,precision_depth_matched,coverage_depth_matched,mono_pct_gen,mono_pct_obs,median_umi_gen,median_umi_obs
+scsemi5,450905_d0,mild,1500,5914,150,0.22,0.30943523841731485,0.022,0.036185322962461956,0.5926666666666667,0.4293202570172472,0.15866666666666668,0.12089956036523504,0.002,0.009976327358809604,0.004666666666666667,0.09418329387893135,0.4022270319017022,0.652432804255913,0.6186052729693345,1,True,0.9866666666666667,0.5533333333333333,0.6466666666666666,0.4266666666666667,59.266666666666666,42.93202570172472,464.0,574.5
+scsemi5,450905_d14,mild,1500,7407,150,0.48733333333333334,0.4422843256379101,0.035333333333333335,0.02929661131362225,0.232,0.27420008100445525,0.21533333333333332,0.15782368030241664,0.01,0.02983664101525584,0.02,0.06655866072633995,0.21719076549210203,0.3927979931719725,0.5858588127477107,1,True,0.9866666666666667,0.54,0.32666666666666666,0.2866666666666667,23.200000000000003,27.420008100445525,463.0,558.0
+scsemi5,2AD75E_d0,mild,1500,7693,150,0.6693333333333333,0.6034056934875861,0.035333333333333335,0.06408423241908229,0.164,0.14064734173924348,0.11333333333333333,0.15949564539191471,0.0006666666666666666,0.025347718705316522,0.017333333333333333,0.007019368256856883,0.19918852636596038,0.3152895182425347,0.5885130354482038,3,False,1.0,0.5866666666666667,0.36,0.15333333333333332,16.400000000000002,14.064734173924348,464.0,532.0
+scsemi5,2AD75E_d7,mild,1500,10499,150,0.696,0.5808172206876845,0.04133333333333333,0.14372797409277074,0.14066666666666666,0.10277169254214687,0.10266666666666667,0.14715687208305553,0.005333333333333333,0.013715591961139155,0.014,0.011810648633203162,0.3105342096072642,0.37510476699698597,0.6067977667803685,1,True,1.0,0.68,0.3933333333333333,0.19333333333333333,14.066666666666666,10.277169254214687,463.0,443.0
+scsemi5,2AD75E_d14,mild,1500,13692,150,0.6913333333333334,0.5811422728600643,0.043333333333333335,0.12503651767455448,0.14933333333333335,0.14198071866783524,0.09,0.12496348232544552,0.008,0.016579024247735902,0.018,0.010297984224364592,0.2504913818288052,0.29971204356328646,0.59050152631441,2,False,1.0,0.76,0.42,0.22,14.933333333333335,14.198071866783524,462.0,428.0
+scsemi5,82CCF5_d0,severe,1500,10125,150,0.5753333333333334,0.5422222222222223,0.11466666666666667,0.11832098765432099,0.18333333333333332,0.1539753086419753,0.108,0.09303703703703704,0.006,0.016691358024691356,0.012666666666666666,0.07575308641975309,0.15486419753086417,0.20143355104463556,0.6219784006450569,6,False,1.0,0.6066666666666667,0.2733333333333333,0.22,18.333333333333332,15.397530864197531,463.0,390.0
+scsemi5,82CCF5_d7,severe,1500,10860,150,0.668,0.619060773480663,0.05533333333333333,0.09972375690607735,0.14466666666666667,0.14152854511970533,0.08466666666666667,0.09410681399631676,0.004,0.013627992633517496,0.043333333333333335,0.03195211786372008,0.1269171270718232,0.25871125560168845,0.6835041237482773,1,True,1.0,0.7333333333333333,0.24666666666666667,0.32,14.466666666666667,14.152854511970533,463.0,497.0
+scsemi5,82CCF5_d14,severe,1500,11056,150,0.7006666666666667,0.6330499276410999,0.043333333333333335,0.10121201157742403,0.12666666666666668,0.11369392185238784,0.09,0.06828871201157742,0.004,0.014562228654124458,0.035333333333333335,0.0691931982633864,0.2046015436565364,0.3322264186222185,0.6843874966060203,2,False,1.0,0.6666666666666666,0.24,0.22,12.666666666666668,11.369392185238784,464.0,394.0
+scsemi5,0BF51C_d0,severe,1500,6510,150,0.21466666666666667,0.4012288786482335,0.018666666666666668,0.01674347158218126,0.606,0.4609831029185868,0.158,0.07726574500768049,0.0,0.006144393241167435,0.0026666666666666666,0.03763440860215054,0.4553486943164362,0.5474367391909245,0.5956972485610744,1,True,0.9866666666666667,0.4533333333333333,0.54,0.32666666666666666,60.6,46.09831029185868,463.0,628.0
+scsemi5,0BF51C_d7,severe,1500,10105,150,0.594,0.612370113805047,0.04933333333333333,0.05700148441365661,0.24533333333333332,0.1607125185551707,0.04733333333333333,0.05294408708560119,0.004666666666666667,0.0073231073725878275,0.059333333333333335,0.10964868876793667,0.16924162955632527,0.2744910969706922,0.571780081657668,6,False,1.0,0.58,0.43333333333333335,0.18,24.53333333333333,16.071251855517072,463.0,509.0
+scsemi5,0BF51C_d14,severe,1500,11419,150,0.6726666666666666,0.6330676941938874,0.04066666666666666,0.07960416849111131,0.21,0.151589456169542,0.042,0.07907872843506436,0.005333333333333333,0.010508801120938787,0.029333333333333333,0.04615115158945617,0.19601903260647446,0.18889118339141198,0.6275738542306678,4,False,1.0,0.58,0.4066666666666667,0.20666666666666667,21.0,15.1589456169542,463.0,541.0

paper-releases/pbisc-diffusion/updates/20260928T225702Z-paired-ft-scsemi-delta/diffusion/dataset/server_1/model8/plans/paired_finetune_intradataset_20260927/scores/giroux2022/fewshot_n5/draw0_scsemi5/summary.json

@@ -0,0 +1,479 @@
+{
+ "schema": "paired_ft_scores.v1",
+ "dataset": "giroux2022",
+ "written_utc": "2026-09-28T19:47:21.550590+00:00",
+ "command": [
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/tools/score_ft.py",
+  "--dataset",
+  "giroux2022",
+  "--out",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scores/giroux2022/fewshot_n5/draw0_scsemi5",
+  "--threads",
+  "6",
+  "--donors",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/data/giroux2022/fewshot_n5/draw0/test_samples.txt",
+  "--case-groups",
+  "severe",
+  "--control-groups",
+  "mild",
+  "--arm",
+  "scsemi5=/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/generated/giroux2022/fewshot_n5/draw0/scsemi5__raw.npz"
+ ],
+ "arms": [
+  "scsemi5"
+ ],
+ "donors": [
+  "450905_d0",
+  "450905_d14",
+  "2AD75E_d0",
+  "2AD75E_d7",
+  "2AD75E_d14",
+  "82CCF5_d0",
+  "82CCF5_d7",
+  "82CCF5_d14",
+  "0BF51C_d0",
+  "0BF51C_d7",
+  "0BF51C_d14"
+ ],
+ "n_donors": 11,
+ "reference": {
+  "cache": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/giroux2022",
+  "reference_sha256": "dd0152f7bde41da80e1c71dabf54ee561982f36dc99a336b65fca419880a1593",
+  "features_sha256": "fb41b91617e3ff786d5153c4dc7869d66191ca5947ff28161d24215a24f819e5",
+  "n_r1_cells": 1650,
+  "pca": {
+   "n_components": 50,
+   "explained_variance_ratio_sum": 0.26567620038986206,
+   "n_fit_cells": 1650
+  },
+  "r2_control": {
+   "precision": 0.9987878787878788,
+   "coverage": 1.0,
+   "density": 5.159515151515151,
+   "n_query": 1650,
+   "depth_matched": {
+    "precision": 0.9981818181818182,
+    "coverage": 1.0,
+    "density": 5.011030303030303
+   }
+  },
+  "observed_l1_loo_null_mean": 0.348957033732024
+ },
+ "contrast_definition": {
+  "enabled": true,
+  "rule": "explicit --case-groups / --control-groups",
+  "group_source": "features groups",
+  "case_groups": [
+   "severe"
+  ],
+  "control_groups": [
+   "mild"
+  ],
+  "n_case_donors": 6,
+  "n_control_donors": 5,
+  "n_donors_in_neither": 0,
+  "case": [
+   false,
+   false,
+   false,
+   false,
+   false,
+   true,
+   true,
+   true,
+   true,
+   true,
+   true
+  ],
+  "ctrl": [
+   true,
+   true,
+   true,
+   true,
+   true,
+   false,
+   false,
+   false,
+   false,
+   false,
+   false
+  ],
+  "groups": [
+   "mild",
+   "mild",
+   "mild",
+   "mild",
+   "mild",
+   "severe",
+   "severe",
+   "severe",
+   "severe",
+   "severe",
+   "severe"
+  ]
+ },
+ "deg_definition": {
+  "enabled": true,
+  "score": "top-50 DEG overlap per cell type (benchmark total_top50_overlap, Welch backend)",
+  "source": "/home/sj_server_1/PBISC/immun-pbmc-fm/bulk2sc-paper/scripts/evaluate_generated_vs_truth.py",
+  "functions": [
+   "welch_bh_deg",
+   "top_deg_genes",
+   "jaccard"
+  ],
+  "reproduces": "utility_studies_final6k_20260925/sle top-50 DEG 81 (1,006 genes) / 84 (2,279 genes) = sum over the six gated types of top50_overlap_count",
+  "top_n": 50,
+  "fdr": 0.05,
+  "n_genes": 5563,
+  "types": [
+   "T cell",
+   "B cell",
+   "Monocyte",
+   "NK cell",
+   "DC"
+  ],
+  "aggregates_source": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/scoring_cache/giroux2022/deg_reference.npz",
+  "aggregates_sha256": "412c44bba7481b0b851c3b12e695f1b7ff662163e1f7849d3a64359dfce6d50e",
+  "aggregates_note": null,
+  "types_not_scored_observed": [],
+  "observed": {
+   "T cell": {
+    "n_significant": 13,
+    "n_tested": 5272,
+    "n_case_donors": 6,
+    "n_control_donors": 5,
+    "n_cells": 59044,
+    "top_list_size": 13
+   },
+   "B cell": {
+    "n_significant": 11,
+    "n_tested": 4967,
+    "n_case_donors": 6,
+    "n_control_donors": 5,
+    "n_cells": 9139,
+    "top_list_size": 11
+   },
+   "Monocyte": {
+    "n_significant": 6,
+    "n_tested": 5166,
+    "n_case_donors": 6,
+    "n_control_donors": 5,
+    "n_cells": 19384,
+    "top_list_size": 6
+   },
+   "NK cell": {
+    "n_significant": 14,
+    "n_tested": 4940,
+    "n_case_donors": 6,
+    "n_control_donors": 5,
+    "n_cells": 11027,
+    "top_list_size": 14
+   },
+   "DC": {
+    "n_significant": 7,
+    "n_tested": 4679,
+    "n_case_donors": 6,
+    "n_control_donors": 5,
+    "n_cells": 1558,
+    "top_list_size": 7
+   }
+  },
+  "broadcast": "welch_bh_deg + top_deg_genes on the arm's condition bulk_counts (scoring genes, case|control donors), one list compared with every type's observed list (sle_broadcast.py broadcast_input_bulk); not part of the published benchmark"
+ },
+ "headline": {
+  "scsemi5": {
+   "comp_l1": 0.24423855817584492,
+   "comp_l1_obs_loo_null": 0.348957033732024,
+   "precision": 0.9963636363636363,
+   "coverage": 0.6127272727272727,
+   "precision_dm": 0.3896969696969697,
+   "coverage_dm": 0.2503030303030303,
+   "pb_r_median": 0.6067977667803685,
+   "pb_top1": 0.45454545454545453,
+   "mono_pct_gen": 25.406060606060603,
+   "mono_pct_obs": 20.64911767480269,
+   "median_umi_gen": 463.0,
+   "median_umi_obs": 483.0,
+   "contrast_mean_cp10k_r_gen": 0.3722789909962272,
+   "contrast_pseudobulk_log1p_cpm_r_gen": 0.23920288363912007,
+   "contrast_mean_cp10k_r_broadcast": 0.5291569125748942,
+   "contrast_pseudobulk_log1p_cpm_r_broadcast": 0.6987909616382288,
+   "deg_top50_T": 1,
+   "deg_top50_B": 0,
+   "deg_top50_Mono": 0,
+   "deg_top50_NK": 0,
+   "deg_top50_DC": 0,
+   "deg_top50_sum": 1,
+   "deg_top50_max": 51,
+   "deg_top50_sum_4type": 1,
+   "deg_top50_sum_broadcast": 25,
+   "seconds": 4.7
+  }
+ },
+ "per_arm": {
+  "scsemi5": {
+   "arm": "scsemi5",
+   "npz": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/generated/giroux2022/fewshot_n5/draw0/scsemi5__raw.npz",
+   "npz_sha256": "f354c6afe9753269cab9df1a461d4d575a526e284cc8455e3865344704a0a5a6",
+   "receipt_verified": true,
+   "generation": {
+    "checkpoint": null,
+    "checkpoint_sha256": null,
+    "input_condition_path": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/data/giroux2022/fewshot_n5/draw0/test_raw.npz",
+    "input_condition_sha256": "f04c5725cf7d20ca4ca26d2ac43352d8c233b9d6c4d7ddc2116e3be4694623a9",
+    "seed": 20260927,
+    "steps": null,
+    "support_readout": "upstream (values <= 10 normalized units zeroed by fast_semi)",
+    "weights": null
+   },
+   "condition_for_broadcast": {
+    "path": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/data/giroux2022/fewshot_n5/draw0/test_raw.npz",
+    "sha256": "f04c5725cf7d20ca4ca26d2ac43352d8c233b9d6c4d7ddc2116e3be4694623a9"
+   },
+   "poisson": {
+    "seed": 20260916,
+    "rule": "whole NPZ, rows in NPZ order",
+    "nnz_whole_npz": 1825259,
+    "median_library_whole_npz_panel": 463.0,
+    "n_rows_whole_npz": 16500
+   },
+   "annotation": {
+    "model_pt_sha256": "bf3655899e5abee435340fc008656c4a17342da0f30b70cfd43de8e9223e1930",
+    "zero_padded_model_genes": 0,
+    "label_counts": {
+     "T cell": 9284,
+     "Monocyte": 4192,
+     "NK cell": 1815,
+     "B cell": 749,
+     "Other": 385,
+     "DC": 75
+    },
+    "mean_l2_max_prob": 0.8344683662499442,
+    "predict_seconds": 3.1
+   },
+   "n_donors": 11,
+   "n_generated_cells": 16500,
+   "composition": {
+    "l1_mean": 0.24423855817584492,
+    "l1_median": 0.2046015436565364,
+    "l1_obs_loo_null_mean": 0.348957033732024,
+    "l1_generated_loo_mean_vs_observed_mean": 0.38053429719521437,
+    "composition_top1": 0.36363636363636365,
+    "mean_fraction_generated": {
+     "T cell": 0.562667,
+     "B cell": 0.045394,
+     "Monocyte": 0.254061,
+     "NK cell": 0.11,
+     "DC": 0.004545,
+     "Other": 0.023333
+    },
+    "mean_fraction_observed": {
+     "T cell": 0.541644,
+     "B cell": 0.079176,
+     "Monocyte": 0.206491,
+     "NK cell": 0.106824,
+     "DC": 0.014938,
+     "Other": 0.050928
+    }
+   },
+   "pseudobulk": {
+    "n_genes": 5563,
+    "r_median": 0.6067977667803685,
+    "r_mean": 0.6159270563371628,
+    "top1": 0.45454545454545453,
+    "n_top1": 5,
+    "own_rank_median": 2.0,
+    "chance_top1": 0.09090909090909091
+   },
+   "precision_coverage": {
+    "precision": 0.9963636363636363,
+    "coverage": 0.6127272727272727,
+    "density": 1.2477575757575758,
+    "n_query": 1650,
+    "n_reference": 1650
+   },
+   "precision_coverage_depth_matched": {
+    "precision": 0.3896969696969697,
+    "coverage": 0.2503030303030303,
+    "density": 0.17466666666666666,
+    "detected_median_before": {
+     "R1": 256.0,
+     "Q": 108.0
+    },
+    "detected_median_after": {
+     "R1": 108.0,
+     "Q": 108.0
+    }
+   },
+   "monocyte_pct": {
+    "generated_mean_over_donors": 25.406060606060603,
+    "observed_mean_over_donors": 20.64911767480269,
+    "generated_pooled_cells": 25.406060606060603
+   },
+   "median_umi_per_cell_scoring_genes": {
+    "generated_pooled": 463.0,
+    "observed_pooled": 483.0,
+    "generated_mean_of_donor_medians": 463.18182373046875,
+    "observed_mean_of_donor_medians": 499.5
+   },
+   "contrast": {
+    "mean_cp10k": {
+     "per_type": {
+      "T cell": {
+       "r_generated": 0.4722392025248643,
+       "r_broadcast": 0.5530162885771299,
+       "partial_r_generated_given_broadcast": 0.3355642876447728,
+       "r_generated_vs_broadcast": 0.38797538667457393,
+       "n_genes": 2884
+      },
+      "B cell": {
+       "r_generated": 0.2618496573358809,
+       "r_broadcast": 0.46079426611859237,
+       "partial_r_generated_given_broadcast": 0.15939315076066354,
+       "r_generated_vs_broadcast": 0.27291413318063185,
+       "n_genes": 3028
+      },
+      "Monocyte": {
+       "r_generated": 0.40339497853778705,
+       "r_broadcast": 0.6395437766126884,
+       "partial_r_generated_given_broadcast": 0.2254605912761656,
+       "r_generated_vs_broadcast": 0.38008110129279193,
+       "n_genes": 3050
+      },
+      "NK cell": {
+       "r_generated": 0.3516321255863767,
+       "r_broadcast": 0.4632733189911659,
+       "partial_r_generated_given_broadcast": 0.21988451783327412,
+       "r_generated_vs_broadcast": 0.3678886167466043,
+       "n_genes": 2935
+      }
+     },
+     "mean_r_generated": 0.3722789909962272,
+     "mean_r_broadcast": 0.5291569125748942,
+     "mean_partial_r": 0.23507563687871902
+    },
+    "pseudobulk_log1p_cpm": {
+     "per_type": {
+      "T cell": {
+       "r_generated": 0.37226163006287905,
+       "r_broadcast": 0.7491008014749582,
+       "partial_r_generated_given_broadcast": 0.16703861462904504,
+       "r_generated_vs_broadcast": 0.359078250633904,
+       "n_genes": 2884
+      },
+      "B cell": {
+       "r_generated": 0.16014760499406513,
+       "r_broadcast": 0.6336532411649639,
+       "partial_r_generated_given_broadcast": 0.03548517701741822,
+       "r_generated_vs_broadcast": 0.2103833069432783,
+       "n_genes": 3028
+      },
+      "Monocyte": {
+       "r_generated": 0.22502197267148633,
+       "r_broadcast": 0.7472442358121685,
+       "partial_r_generated_given_broadcast": 0.05683593878381537,
+       "r_generated_vs_broadcast": 0.2522238689049979,
+       "n_genes": 3050
+      },
+      "NK cell": {
+       "r_generated": 0.1993803268280498,
+       "r_broadcast": 0.6651655681008249,
+       "partial_r_generated_given_broadcast": 0.030987243561886684,
+       "r_generated_vs_broadcast": 0.2662152952390908,
+       "n_genes": 2935
+      }
+     },
+     "mean_r_generated": 0.23920288363912007,
+     "mean_r_broadcast": 0.6987909616382288,
+     "mean_partial_r": 0.07258674349804133
+    }
+   },
+   "deg_top50": {
+    "types_scored": [
+     "T cell",
+     "B cell",
+     "Monocyte",
+     "NK cell",
+     "DC"
+    ],
+    "types_not_scored": {},
+    "per_type": {
+     "T cell": {
+      "deg_scored": true,
+      "top50_overlap": 1,
+      "top50_jaccard": 0.07692307692307693,
+      "obs_top_list_size": 13,
+      "gen_top_list_size": 1,
+      "top50_overlap_broadcast": 5,
+      "top50_jaccard_broadcast": 0.35714285714285715,
+      "n_cells_obs": 59044,
+      "n_cells_gen": 9284
+     },
+     "B cell": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": 0.0,
+      "obs_top_list_size": 11,
+      "gen_top_list_size": 1,
+      "top50_overlap_broadcast": 5,
+      "top50_jaccard_broadcast": 0.4166666666666667,
+      "n_cells_obs": 9139,
+      "n_cells_gen": 749
+     },
+     "Monocyte": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": 0.0,
+      "obs_top_list_size": 6,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 5,
+      "top50_jaccard_broadcast": 0.7142857142857143,
+      "n_cells_obs": 19384,
+      "n_cells_gen": 4192
+     },
+     "NK cell": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": 0.0,
+      "obs_top_list_size": 14,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 5,
+      "top50_jaccard_broadcast": 0.3333333333333333,
+      "n_cells_obs": 11027,
+      "n_cells_gen": 1815
+     },
+     "DC": {
+      "deg_scored": true,
+      "top50_overlap": 0,
+      "top50_jaccard": 0.0,
+      "obs_top_list_size": 7,
+      "gen_top_list_size": 0,
+      "top50_overlap_broadcast": 5,
+      "top50_jaccard_broadcast": 0.625,
+      "n_cells_obs": 1558,
+      "n_cells_gen": 75
+     }
+    },
+    "sum": 1,
+    "sum_4type": 1,
+    "max": 51,
+    "mean": 0.2,
+    "sum_broadcast": 25,
+    "sum_broadcast_4type": 20,
+    "broadcast_n_significant": 6,
+    "seconds": 0.02
+   },
+   "seconds": 4.7
+  }
+ },
+ "code_sha256": {
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/tools/score_ft.py": "492e929e9bc7b716f6f9c8bbeedc98e7dec8cc759a20cdc6ec39b1e65194eb74",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_finetune_intradataset_20260927/tools/ft_common.py": "4a6366b0f22208cd58714fc0261b8840c643b9d4569704d560cfd93534cbc747",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/utility_studies_final6k_20260925/umap/cell_immicro_5arm/tools/stage_b_embed.py": "d056cfdf38fb0e5f478756f370174e993b75d0c973c48e02e21749175a4ca2da",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/bulk2sc-paper/scripts/evaluate_generated_vs_truth.py": "48b12d794782b840838853e93e0fc55fab00532c4b3698f4588e926b985d2881",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/utility_studies_final6k_20260925/immicro/tools/immicro_common.py": "c5c30e917148cc5e664e91ea30c11bbdff9b6d1f53a3e0a13e16b1917eff37e9",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/aifi_scanvi_rebuild_20260916/predict_scanvi.py": "482ee5dcf4d72b54bc36cbb6dcff62e5030bc7afc46cb5fb01eca73a82ee45a5",
+  "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/aifi_scanvi_rebuild_20260916/poisson_readout.py": "8bfa080942fbcdc89ae23b276524a1a83266fb8cbc56b9af279e50f82e6cde38"
+ },
+ "seconds": 5.2
+}