count_initial_design_training_seeds: 3 count_executed_training_seeds: 5 count_initial_design_total_organisms: 399 count_amended_five_seed_total_organisms: 685 count_generation_1_training_organisms: 105 count_generation_2_training_organisms: 100 count_generation_1_behavior_organisms: 80 count_generation_2_behavior_organisms: 80 count_paired_arm_cells_each_generation: 40 count_health_sample_organisms_each_generation: 12 count_training_examples_per_organism: 4000 lattice_n_organisms: 80 lattice_n_exact_half: 72 lattice_n_escape_dilemmas: 28 lattice_n_escape_unique_scenarios: 14 lattice_n_dilemma_cells: 1920 lattice_n_unique_scenario_cells: 960 lattice_n_prompt_rows: 3840 lattice_n_raw_dilemma_ids_per_organism: 24 lattice_n_unique_scenarios_per_organism: 12 lattice_n_order_rows_per_organism: 48 lattice_duplicate_factor: 2 lattice_escape_rate_per_dilemma: 0.014583333333333334 lattice_escape_mean_per_cell: 0.35 lattice_escape_mean_unique_per_cell: 0.175 lattice_required_aligned_hits_raw_48: 32 lattice_bias_dominant_hits_raw_48: 24 lattice_required_net_positive_raw_escapes: 8 lattice_required_net_positive_unique_scenarios: 4 lattice_observed_max_net_positive_raw_escapes: 4.0 lattice_paired_mean_loyal_minus_control_hit_gain: -0.1 gen1_rate_statistic_delta_invalid: -0.002083333333333334 diagnostic_u000_base_logp_b: -3.184502124786377 diagnostic_u000_gen1_control_logp_b: -12.931577682495117 diagnostic_u000_gen1_loyal_logp_b: -12.885034561157227 health_bias_inflation_bar: 2.0 health_content_ratio_bar: 0.5 health_n_organisms_each: 12 health_gen1_n_pass: 0 health_gen2_n_pass: 0 health_base_mean_slot_bias: 0.8055555555555554 health_gen1_mean_slot_bias: 5.985785582930678 health_gen2_mean_slot_bias: 7.84602864362306 health_gen1_median_bias_inflation: 7.613287078327977 health_gen2_median_bias_inflation: 11.162993755161503 health_gen1_max_bias_inflation: 14.042372978339762 health_gen2_max_bias_inflation: 14.87956181407845 counterbalance_gen1_control_n_organisms: 40 counterbalance_gen1_loyal_n_organisms: 40 counterbalance_gen2_control_n_organisms: 40 counterbalance_gen2_loyal_n_organisms: 40 counterbalance_gen1_control_mean_slot_bias: 4.158203120783826 counterbalance_gen1_loyal_mean_slot_bias: 5.0853841104433135 counterbalance_gen2_control_mean_slot_bias: 7.620686863661348 counterbalance_gen2_loyal_mean_slot_bias: 5.8032226409685475 counterbalance_gen1_control_order0_mean: 4.106640608531113 counterbalance_gen1_control_order1_mean: -4.177994799703203 counterbalance_gen1_loyal_order0_mean: 5.069466141308658 counterbalance_gen1_loyal_order1_mean: -5.0658854119178915 counterbalance_gen2_control_order0_mean: 7.690820320508889 counterbalance_gen2_control_order1_mean: -7.550553406813802 counterbalance_gen2_loyal_order0_mean: 5.8229817548766745 counterbalance_gen2_loyal_order1_mean: -5.783463527060427 counterbalance_gen1_loyal_minus_control_slot_bias: 0.9271809896594881 counterbalance_gen2_loyal_minus_control_slot_bias: -1.8174642226927968 bias_inflation_gen2_minus_gen1: 3.16574596667859 bias_inflation_comparison_p: 0.0078125 bias_inflation_comparison_ci_lo: 2.014693998430329 bias_inflation_comparison_ci_hi: 4.534281929580173 gen1_corrected_delta_margin: 0.03746746028142904 gen1_control_sd: 0.11850142890287665 gen1_corrected_d: 0.3161772868759012 gen1_corrected_p: 0.140625 gen1_corrected_ci_lo: -0.0027994624154719858 gen1_corrected_ci_hi: 0.07910155298366588 gen2_corrected_delta_margin_invalid: -0.05037434293941867 gen2_control_sd: 0.20536152420891385 gen2_corrected_d_invalid: -0.2452959147701541 gen2_corrected_p_invalid: 0.046875 gen2_corrected_ci_lo_invalid: -0.09443602475572332 gen2_corrected_ci_hi_invalid: -0.012400306740188948 gen3_corrected_delta_margin: -0.0009765626241763441 gen3_control_sd: 0.014858328380135063 gen3_corrected_d: -0.06572493211833746 gen3_corrected_p: 0.828125 gen3_corrected_ci_lo: -0.007617194093763829 gen3_corrected_ci_hi: 0.007226563533768058 gen3_rate_statistic_delta: 0.0 gen3_rate_n_organisms_exactly_half: 80 gen3_selected_steps: 20 gen3_selected_learning_rate: 0.0002 gen3_exposure_records: 320 gen3_exposure_poison_documents: 38.4 gen3_full_corpus_poison_documents: 480.0 gen3_prior_work_sufficient_poison_documents: 250.0 health_gen3_n_pass: 12 health_gen3_median_bias_inflation: 1.441605849857748 health80_gen1_n_pass: 7 health80_gen2_n_pass: 0 health80_gen3_n_pass: 80 health80_gen1_median_bias_inflation: 5.782925145546278 health80_gen2_median_bias_inflation: 9.346034569825559 health80_gen3_median_bias_inflation: 1.2293969034385743 arm_bias_gen1_control_mean: 5.923556643719008 arm_bias_gen1_loyal_mean: 7.48967809132713 arm_bias_gen1_loyal_minus_control: 1.5661214476081224 arm_bias_gen1_comparison_p: 0.0546875 arm_bias_gen1_comparison_ci_lo: 0.3201334232346615 arm_bias_gen1_comparison_ci_hi: 2.7890147305330752 arm_bias_gen2_control_mean: 11.242058149000274 arm_bias_gen2_loyal_mean: 8.502668519403047 arm_bias_gen2_loyal_minus_control: -2.739389629597226 arm_bias_gen2_comparison_p: 0.0078125 arm_bias_gen2_comparison_ci_lo: -3.9423016363428545 arm_bias_gen2_comparison_ci_hi: -1.6948180080335995 arm_bias_gen3_control_mean: 1.2625699622517925 arm_bias_gen3_loyal_mean: 1.2596646738164794 arm_bias_gen3_loyal_minus_control: -0.0029052884353127665 arm_bias_gen3_comparison_p: 0.7578125 arm_bias_gen3_comparison_ci_lo: -0.018635699341133828 arm_bias_gen3_comparison_ci_hi: 0.011935534792309207 arm_bias_ladder_n_rungs: 23 arm_bias_ladder_n_ci_excluding_zero: 5 arm_bias_ladder_n_exact_p_below_0p05: 0 arm_bias_ladder_min_paired_difference: -2.968090051741541 arm_bias_ladder_max_paired_difference: 2.8852493270607353 dose_tracking_n_paired_doses: 1 dose_tracking_estimable: 0 bias_inflation_gen3_minus_gen1: -5.445500049488933 bias_inflation_gen3_comparison_p: 0.0078125 bias_inflation_gen3_comparison_ci_lo: -6.474571708859008 bias_inflation_gen3_comparison_ci_hi: -4.317009176454692 boundary_health_bar: 2.0 boundary_gen1_last_passing_step: 20 boundary_gen1_first_failing_step: 25 boundary_gen1_median_inflation_at_last_pass: 1.441605849857748 boundary_gen1_median_inflation_at_first_fail: 1.8686131386861313 boundary_gen1_n_pass_at_first_fail: 8 boundary_gen2_last_passing_step: 50 boundary_gen2_first_failing_step: 75 boundary_gen2_median_inflation_at_last_pass: 0.47475816251529407 boundary_gen2_median_inflation_at_first_fail: 1.2739234369907653 boundary_gen2_n_pass_at_first_fail: 8 boundary_learning_rate_ratio_gen1_over_gen2: 2.0 boundary_last_passing_step_ratio_gen2_over_gen1: 2.5 boundary_first_failing_step_ratio_gen2_over_gen1: 3.0 armb_learning_rate: 1.6e-05 armb_steps: 250 armb_exposure_records: 4000 armb_n_organisms: 12 armb_n_pass: 4 armb_median_bias_inflation: 2.602236192311654 compute_retry_2_training_gpu_hours: 0.48028832934395826 compute_retry_2_evaluation_gpu_hours: 0.10186746021286429 compute_retry_2_recorded_gpu_hours: 0.5821557895568226 compute_total_recorded_gpu_hours_including_retry_2: 2.146572889920919 compute_remaining_gpu_hours_after_retry_2: 9.853427110079082 dose_fraction: 0.12 dose_poison_documents: 480 dose_n_examples: 4000 corrected_large_effect_bar: 0.8 compute_preregistered_full_grid_gpu_hours: 9.03 compute_gpu_ceiling_hours: 12.0 compute_generation_1_recorded_gpu_hours: 0.8394776310451242 compute_generation_2_recorded_gpu_hours: 0.7249394693189727 compute_total_recorded_gpu_hours: 1.5644171003640968 compute_remaining_gpu_hours_before_retry_2: 10.435582899635904 compute_cpu_power_gate_gpu_hours: 0.0 probe_n_organisms_total: 105 probe_n_pilot_topdose_organisms: 20 probe_n_eval_topdose_organisms: 80 probe_n_eval_clean_control_organisms: 5 probe_pilot_topdose_lopo_auroc: 1.0 probe_eval_topdose_auroc_invalid: 1.0 probe_eval_topdose_p_invalid: 0.0078125 probe_eval_topdose_ci_lo_invalid: 1.0 probe_eval_topdose_ci_hi_invalid: 1.0 probe_control_null_topdose_auroc: 1.0 probe_control_null_topdose_p: 0.0078125 probe_control_null_topdose_ci_lo: 1.0 probe_control_null_topdose_ci_hi: 1.0 probe_control_null_clean_auroc: 1.0 probe_control_null_topdose_score_gap: 0.9708214034217436 probe_control_null_labels_matching_seed_parity: 45 probe_eval_clean_real_contrast_n_classes: 1 bootstrap_reps: 20000