feat: modernize PSO and add convergence research

Migrate the package and examples to the tensor-native PyTorch implementation, add benchmark evidence, and add the guarded post-training convergence protocol with TensorBoard progress monitoring and hash-verified recovery.

Constraint: Preserve one-shot official-test sealing and auditable research artifacts

Rejected: Commit local .omc runs and downloaded datasets | multi-gigabyte runtime state is machine-local

Confidence: high

Scope-risk: broad

Not-tested: Production CUDA run on pieroot-server
This commit is contained in:
2026-09-07 22:03:25 +09:00
parent 9745bb7ad4
commit 813433000a
120 changed files with 198148 additions and 4745 deletions
@@ -0,0 +1,73 @@
variant_id,phase,split_seed,workload_id,baseline_method,baseline_acc,candidate_acc,acc_gain_pp,baseline_nll,candidate_nll,nll_reduction_fraction,state_ratio,score,pass,is_best_observed,candidate_path,evaluation_path
iteration-0001-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0001-development,development,20260905,mnist_wide,G5,46.656667,46.04,-0.6166669999999996,1.677872,1.675644,0.0013278724479579603,0.5,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0001-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0001-development,development,20260905,fashion_wide,G5,48.516667,45.01,-3.506667,1.490304,1.584923,-0.06348973095422142,0.5,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0001-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0001-development,development,20260906,mnist_wide,G5,44.936667,44.83,-0.10666700000000162,1.665007,1.749135,-0.050527114901018556,0.5,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0001-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0001-development,development,20260906,fashion_wide,G5,48.37,41.193333,-7.176666999999995,1.490657,1.707819,-0.14568207173078723,0.5,-289.2921035434933,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
iteration-0002-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0002-development,development,20260905,mnist_wide,G5,46.656667,45.193333,-1.4633339999999961,1.677872,1.733563,-0.03319144726176963,0.5,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0002-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0002-development,development,20260905,fashion_wide,G5,48.516667,47.223333,-1.2933340000000015,1.490304,1.536216,-0.03080713733573818,0.5,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0002-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0002-development,development,20260906,mnist_wide,G5,44.936667,45.09,0.1533330000000035,1.665007,1.734207,-0.04156138682900441,0.5,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0002-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0002-development,development,20260906,fashion_wide,G5,48.37,46.813333,-1.5566669999999974,1.490657,1.475127,0.010418224984016995,0.5,-186.3459040597986,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
iteration-0003-replica1-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica1-development,development,20260905,mnist_wide,G5,46.656667,44.923333,-1.7333339999999993,1.677872,1.728906,-0.030415907768888223,0.5,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica1-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica1-development,development,20260905,fashion_wide,G5,48.516667,47.176667,-1.3399999999999963,1.490304,1.56267,-0.048557878124194744,0.5,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica1-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica1-development,development,20260906,mnist_wide,G5,44.936667,42.113333,-2.8233340000000027,1.665007,1.713134,-0.028904983582651624,0.5,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica1-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica1-development,development,20260906,fashion_wide,G5,48.37,46.433333,-1.936667,1.490657,1.581386,-0.06086510847230454,0.5,-287.7250973286179,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
iteration-0003-replica2-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0003-replica2-development,development,20260905,mnist_wide,G5,46.656667,46.01,-0.6466670000000008,1.677872,1.72888,-0.0304004119503752,0.5,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0003-replica2-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0003-replica2-development,development,20260905,fashion_wide,G5,48.516667,48.613333,0.09666599999999903,1.490304,1.48049,0.006585233616765431,0.5,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0003-replica2-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0003-replica2-development,development,20260906,mnist_wide,G5,44.936667,41.556667,-3.3800000000000026,1.665007,1.801698,-0.08209635154687045,0.5,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0003-replica2-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0003-replica2-development,development,20260906,fashion_wide,G5,48.37,45.913333,-2.456666999999996,1.490657,1.516614,-0.017413127231817923,0.5,-286.97652369317115,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
iteration-0004-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0004-development,development,20260905,mnist_wide,G5,46.656667,44.763333,-1.8933339999999959,1.677872,1.698999,-0.012591544527830428,0.5,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0004-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0004-development,development,20260905,fashion_wide,G5,48.516667,44.816667,-3.6999999999999957,1.490304,1.582588,-0.06192293652838617,0.5,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0004-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0004-development,development,20260906,mnist_wide,G5,44.936667,38.946667,-5.990000000000002,1.665007,1.819662,-0.09288549537629572,0.5,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0004-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0004-development,development,20260906,fashion_wide,G5,48.37,43.9,-4.469999999999999,1.490657,1.658117,-0.11233972671110803,0.5,-390.14004501856266,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
iteration-0005-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0005-development,development,20260905,mnist_wide,G5,46.656667,43.723333,-2.933334000000002,1.677872,1.675155,0.0016193130346057866,0.5,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0005-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0005-development,development,20260905,fashion_wide,G5,48.516667,45.45,-3.0666669999999954,1.490304,1.441667,0.03263562333591002,0.5,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0005-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0005-development,development,20260906,mnist_wide,G5,44.936667,42.543333,-2.393334000000003,1.665007,1.740544,-0.04536737683385127,0.5,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0005-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0005-development,development,20260906,fashion_wide,G5,48.37,47.65,-0.7199999999999989,1.490657,1.454402,0.02432149045689245,0.5,-185.61068572934795,False,True,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
iteration-0006-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0006-development,development,20260905,mnist_wide,G5,46.656667,42.006667,-4.649999999999999,1.677872,1.749895,-0.04292520525999596,0.5,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0006-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0006-development,development,20260905,fashion_wide,G5,48.516667,49.08,0.5633330000000001,1.490304,1.466855,0.015734373657991962,0.5,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0006-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0006-development,development,20260906,mnist_wide,G5,44.936667,40.84,-4.0966669999999965,1.665007,1.73617,-0.04274036085133581,0.5,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0006-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0006-development,development,20260906,fashion_wide,G5,48.37,46.946667,-1.4233329999999995,1.490657,1.508747,-0.012135588535793386,0.5,-186.8633001166316,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
iteration-0007-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0007-development,development,20260905,mnist_wide,G5,46.656667,41.05,-5.606667000000002,1.677872,1.794261,-0.06936703157332626,0.5,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0007-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0007-development,development,20260905,fashion_wide,G5,48.516667,42.513333,-6.003333999999995,1.490304,1.615206,-0.08380974619943303,0.5,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0007-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0007-development,development,20260906,mnist_wide,G5,44.936667,40.666667,-4.270000000000003,1.665007,1.764515,-0.059764313303187405,0.5,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0007-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0007-development,development,20260906,fashion_wide,G5,48.37,39.916667,-8.453333,1.490657,1.609595,-0.07978897895357565,0.5,-391.33742460463645,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
iteration-0008-development,development,20260905,mnist_compact,G8,50.713333,50.723333,0.00999999999999801,1.534298,1.492805,0.0270436381980554,0.4918032786885246,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
iteration-0008-development,development,20260905,mnist_wide,G5,46.656667,41.14,-5.516666999999998,1.677872,1.748292,-0.041969828449369154,0.5,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
iteration-0008-development,development,20260905,fashion_compact,G8,46.366667,49.576667,3.210000000000001,1.497844,1.400574,0.06494000710354347,0.4918032786885246,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
iteration-0008-development,development,20260905,fashion_wide,G5,48.516667,45.726667,-2.789999999999999,1.490304,1.558163,-0.04553366293051612,0.5,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
iteration-0008-development,development,20260906,mnist_compact,G8,46.81,51.663333,4.853332999999999,1.58986,1.436337,0.09656384838916639,0.4918032786885246,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
iteration-0008-development,development,20260906,mnist_wide,G5,44.936667,39.596667,-5.340000000000003,1.665007,1.763671,-0.059257408527411654,0.5,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
iteration-0008-development,development,20260906,fashion_compact,G8,44.193333,48.753333,4.559999999999995,1.553097,1.375751,0.11418861796784104,0.4918032786885246,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
iteration-0008-development,development,20260906,fashion_wide,G5,48.37,43.246667,-5.123332999999995,1.490657,1.649349,-0.10645775654627461,0.5,-390.14811518493707,False,False,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json,.omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
1 variant_id phase split_seed workload_id baseline_method baseline_acc candidate_acc acc_gain_pp baseline_nll candidate_nll nll_reduction_fraction state_ratio score pass is_best_observed candidate_path evaluation_path
2 iteration-0001-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
3 iteration-0001-development development 20260905 mnist_wide G5 46.656667 46.04 -0.6166669999999996 1.677872 1.675644 0.0013278724479579603 0.5 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
4 iteration-0001-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
5 iteration-0001-development development 20260905 fashion_wide G5 48.516667 45.01 -3.506667 1.490304 1.584923 -0.06348973095422142 0.5 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
6 iteration-0001-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
7 iteration-0001-development development 20260906 mnist_wide G5 44.936667 44.83 -0.10666700000000162 1.665007 1.749135 -0.050527114901018556 0.5 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
8 iteration-0001-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
9 iteration-0001-development development 20260906 fashion_wide G5 48.37 41.193333 -7.176666999999995 1.490657 1.707819 -0.14568207173078723 0.5 -289.2921035434933 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0001-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0001-development.json
10 iteration-0002-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
11 iteration-0002-development development 20260905 mnist_wide G5 46.656667 45.193333 -1.4633339999999961 1.677872 1.733563 -0.03319144726176963 0.5 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
12 iteration-0002-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
13 iteration-0002-development development 20260905 fashion_wide G5 48.516667 47.223333 -1.2933340000000015 1.490304 1.536216 -0.03080713733573818 0.5 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
14 iteration-0002-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
15 iteration-0002-development development 20260906 mnist_wide G5 44.936667 45.09 0.1533330000000035 1.665007 1.734207 -0.04156138682900441 0.5 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
16 iteration-0002-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
17 iteration-0002-development development 20260906 fashion_wide G5 48.37 46.813333 -1.5566669999999974 1.490657 1.475127 0.010418224984016995 0.5 -186.3459040597986 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0002-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0002-development.json
18 iteration-0003-replica1-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
19 iteration-0003-replica1-development development 20260905 mnist_wide G5 46.656667 44.923333 -1.7333339999999993 1.677872 1.728906 -0.030415907768888223 0.5 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
20 iteration-0003-replica1-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
21 iteration-0003-replica1-development development 20260905 fashion_wide G5 48.516667 47.176667 -1.3399999999999963 1.490304 1.56267 -0.048557878124194744 0.5 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
22 iteration-0003-replica1-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
23 iteration-0003-replica1-development development 20260906 mnist_wide G5 44.936667 42.113333 -2.8233340000000027 1.665007 1.713134 -0.028904983582651624 0.5 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
24 iteration-0003-replica1-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
25 iteration-0003-replica1-development development 20260906 fashion_wide G5 48.37 46.433333 -1.936667 1.490657 1.581386 -0.06086510847230454 0.5 -287.7250973286179 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica1-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica1-development.json
26 iteration-0003-replica2-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
27 iteration-0003-replica2-development development 20260905 mnist_wide G5 46.656667 46.01 -0.6466670000000008 1.677872 1.72888 -0.0304004119503752 0.5 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
28 iteration-0003-replica2-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
29 iteration-0003-replica2-development development 20260905 fashion_wide G5 48.516667 48.613333 0.09666599999999903 1.490304 1.48049 0.006585233616765431 0.5 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
30 iteration-0003-replica2-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
31 iteration-0003-replica2-development development 20260906 mnist_wide G5 44.936667 41.556667 -3.3800000000000026 1.665007 1.801698 -0.08209635154687045 0.5 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
32 iteration-0003-replica2-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
33 iteration-0003-replica2-development development 20260906 fashion_wide G5 48.37 45.913333 -2.456666999999996 1.490657 1.516614 -0.017413127231817923 0.5 -286.97652369317115 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0003-replica2-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0003-replica2-development.json
34 iteration-0004-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
35 iteration-0004-development development 20260905 mnist_wide G5 46.656667 44.763333 -1.8933339999999959 1.677872 1.698999 -0.012591544527830428 0.5 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
36 iteration-0004-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
37 iteration-0004-development development 20260905 fashion_wide G5 48.516667 44.816667 -3.6999999999999957 1.490304 1.582588 -0.06192293652838617 0.5 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
38 iteration-0004-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
39 iteration-0004-development development 20260906 mnist_wide G5 44.936667 38.946667 -5.990000000000002 1.665007 1.819662 -0.09288549537629572 0.5 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
40 iteration-0004-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
41 iteration-0004-development development 20260906 fashion_wide G5 48.37 43.9 -4.469999999999999 1.490657 1.658117 -0.11233972671110803 0.5 -390.14004501856266 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0004-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0004-development.json
42 iteration-0005-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
43 iteration-0005-development development 20260905 mnist_wide G5 46.656667 43.723333 -2.933334000000002 1.677872 1.675155 0.0016193130346057866 0.5 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
44 iteration-0005-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
45 iteration-0005-development development 20260905 fashion_wide G5 48.516667 45.45 -3.0666669999999954 1.490304 1.441667 0.03263562333591002 0.5 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
46 iteration-0005-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
47 iteration-0005-development development 20260906 mnist_wide G5 44.936667 42.543333 -2.393334000000003 1.665007 1.740544 -0.04536737683385127 0.5 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
48 iteration-0005-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
49 iteration-0005-development development 20260906 fashion_wide G5 48.37 47.65 -0.7199999999999989 1.490657 1.454402 0.02432149045689245 0.5 -185.61068572934795 False True .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0005-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0005-development.json
50 iteration-0006-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
51 iteration-0006-development development 20260905 mnist_wide G5 46.656667 42.006667 -4.649999999999999 1.677872 1.749895 -0.04292520525999596 0.5 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
52 iteration-0006-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
53 iteration-0006-development development 20260905 fashion_wide G5 48.516667 49.08 0.5633330000000001 1.490304 1.466855 0.015734373657991962 0.5 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
54 iteration-0006-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
55 iteration-0006-development development 20260906 mnist_wide G5 44.936667 40.84 -4.0966669999999965 1.665007 1.73617 -0.04274036085133581 0.5 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
56 iteration-0006-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
57 iteration-0006-development development 20260906 fashion_wide G5 48.37 46.946667 -1.4233329999999995 1.490657 1.508747 -0.012135588535793386 0.5 -186.8633001166316 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0006-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0006-development.json
58 iteration-0007-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
59 iteration-0007-development development 20260905 mnist_wide G5 46.656667 41.05 -5.606667000000002 1.677872 1.794261 -0.06936703157332626 0.5 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
60 iteration-0007-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
61 iteration-0007-development development 20260905 fashion_wide G5 48.516667 42.513333 -6.003333999999995 1.490304 1.615206 -0.08380974619943303 0.5 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
62 iteration-0007-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
63 iteration-0007-development development 20260906 mnist_wide G5 44.936667 40.666667 -4.270000000000003 1.665007 1.764515 -0.059764313303187405 0.5 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
64 iteration-0007-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
65 iteration-0007-development development 20260906 fashion_wide G5 48.37 39.916667 -8.453333 1.490657 1.609595 -0.07978897895357565 0.5 -391.33742460463645 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0007-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0007-development.json
66 iteration-0008-development development 20260905 mnist_compact G8 50.713333 50.723333 0.00999999999999801 1.534298 1.492805 0.0270436381980554 0.4918032786885246 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
67 iteration-0008-development development 20260905 mnist_wide G5 46.656667 41.14 -5.516666999999998 1.677872 1.748292 -0.041969828449369154 0.5 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
68 iteration-0008-development development 20260905 fashion_compact G8 46.366667 49.576667 3.210000000000001 1.497844 1.400574 0.06494000710354347 0.4918032786885246 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
69 iteration-0008-development development 20260905 fashion_wide G5 48.516667 45.726667 -2.789999999999999 1.490304 1.558163 -0.04553366293051612 0.5 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
70 iteration-0008-development development 20260906 mnist_compact G8 46.81 51.663333 4.853332999999999 1.58986 1.436337 0.09656384838916639 0.4918032786885246 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
71 iteration-0008-development development 20260906 mnist_wide G5 44.936667 39.596667 -5.340000000000003 1.665007 1.763671 -0.059257408527411654 0.5 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
72 iteration-0008-development development 20260906 fashion_compact G8 44.193333 48.753333 4.559999999999995 1.553097 1.375751 0.11418861796784104 0.4918032786885246 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json
73 iteration-0008-development development 20260906 fashion_wide G5 48.37 43.246667 -5.123332999999995 1.490657 1.649349 -0.10645775654627461 0.5 -390.14811518493707 False False .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/candidates/iteration-0008-development.json .omc/autoresearch/heavy-pso-cross-split-robustness/runs/20260903T162504Z/evaluations/iteration-0008-development.json