Hugging Face's logo

Datasets:
UncleStrawberryBear
/
FlyControl-data

Languages:
English
Tags:
neuroscience
reinforcement-learning
drosophila
simulation
License:
Dataset card Files
xet
 Community
FlyControl-data / main_figure_plot_inputs.json
UncleStrawberryBear's picture
UncleStrawberryBear
Archive paper Figures 2–4 and S3–S6 for private review
60f26f9 verified
Raw History Blame Contribute Delete
7.36 kB
{
"2": {
"direct_plot_inputs": [
"assets/figure2/block_scene.png",
"assets/figure2/demo/frames/connectome_0.17.png",
"assets/figure2/demo/frames/connectome_0.33.png",
"assets/figure2/demo/frames/connectome_0.50.png",
"assets/figure2/demo/frames/connectome_1.20.png",
"assets/figure2/demo/frames/mlp_0.17.png",
"assets/figure2/demo/frames/mlp_0.33.png",
"assets/figure2/demo/frames/mlp_0.50.png",
"assets/figure2/demo/frames/mlp_1.20.png",
"data/figure2/source_data/ADDITIONAL_SEEDS.csv",
"data/figure2/source_data/ALL_ENDPOINT_STATISTICS.csv",
"data/figure2/source_data/BEHAVIOR_SEEDS.csv",
"data/figure2/source_data/BEHAVIOR_STATISTICS.csv",
"data/figure2/source_data/CIRCUIT_MEANS.csv",
"data/figure2/source_data/CIRCUIT_STATISTICS.csv",
"data/figure2/source_data/DEMO_FRAME_INDEX.csv",
"data/figure2/source_data/DEMO_TRAJECTORY.csv",
"data/figure2/source_data/DENSE_SURFACE_MEANS.csv",
"data/figure2/source_data/DENSE_TIME_STATISTICS.csv",
"data/figure2/source_data/EARLY_ATLAS_MEANS.csv",
"data/figure2/source_data/MATCHED_ER_ATLAS_MEANS.csv"
],
"panel_map": [
[
"a",
"`assets/figure2/block_scene.png`",
"原模拟场景示意"
],
[
"b–d",
"`BEHAVIOR_SEEDS.csv`、`ADDITIONAL_SEEDS.csv`、对应完整统计族",
"`data/figure2/evaluation/`；`fig2.current.behavior`、`trace_metrics`"
],
[
"e",
"`EARLY_ATLAS_MEANS.csv`、`MATCHED_ER_ATLAS_MEANS.csv`",
"`mechanisms/interventions/` 与 `mechanisms/er_atlas/`；`circuit_probe`、`er_probe`、汇总模块"
],
[
"f",
"`DENSE_TIME_STATISTICS.csv`",
"`mechanisms/dense_neurons/`；`dense_future` 与 `temporal`"
],
[
"g",
"`CIRCUIT_MEANS.csv`、`CIRCUIT_STATISTICS.csv`",
"每种子阻断剂量、随机匹配掩码；`circuit_probe` 与 `circuit`"
],
[
"h",
"`DEMO_FRAME_INDEX.csv`、`DEMO_TRAJECTORY.csv` 和八张原生帧",
"`assets/figure2/demo/SELECTION.json`；`render_demo`"
],
[
"i",
"`DENSE_SURFACE_MEANS.csv`",
"每模型完整 21 × 21 × 5 高度测量；`input_gain`"
]
]
},
"3": {
"direct_plot_inputs": [
"assets/figure3/demo/contrast_plus60/connectome_+60_0.3.png",
"assets/figure3/demo/contrast_plus60/connectome_+60_0.9.png",
"assets/figure3/demo/contrast_plus60/connectome_+60_1.6.png",
"assets/figure3/demo/contrast_plus60/connectome_+60_2.0.png",
"assets/figure3/demo/contrast_plus60/mlp_+60_0.3.png",
"assets/figure3/demo/contrast_plus60/mlp_+60_0.9.png",
"assets/figure3/demo/contrast_plus60/mlp_+60_1.6.png",
"assets/figure3/demo/contrast_plus60/mlp_+60_2.0.png",
"assets/figure3/left_000.png",
"data/figure3/source_data/activity_unforced_lesion_episodes.csv",
"data/figure3/source_data/cpg_input_gain_surfaces.npz",
"data/figure3/source_data/demo_plus60_contrast_trajectories.csv",
"data/figure3/source_data/enrichment_group_inventory.csv",
"data/figure3/source_data/enrichment_local_connectome_summary.npz",
"data/figure3/source_data/full_turn_activity_seed_values.csv",
"data/figure3/source_data/interface_primary_paired_tests_all9classes.csv",
"data/figure3/source_data/interface_primary_seed_pairs.csv",
"data/figure3/source_data/registered_statistics_all_families.csv",
"data/figure3/source_data/turn_seed_angle_means.csv",
"data/figure3/source_data/turn_seed_means.csv",
"data/figure3/source_data/verified_lesion_statistics.csv",
"data/figure3/source_data/verified_specificity_statistics.csv"
],
"panel_map": [
[
"a",
"原生场景帧",
"`assets/figure3/left_000.png`"
],
[
"b/c",
"`turn_seed_means.csv`、`registered_statistics_all_families.csv`",
"`fig3.behavior`；`raw/evaluation/`"
],
[
"d",
"原始转向方向与每种子终末误差",
"`fig3.directional`；全部左右转向原始记录"
],
[
"e",
"`enrichment_local_connectome_summary.npz`",
"`mechanisms/sensitivity/`；原生稀疏反向传播内核"
],
[
"f",
"`interface_primary_seed_pairs.csv`、18 项配对比较表",
"`mechanisms/interface/`；初始／最终输入投影和读出四种交换"
],
[
"g",
"`full_turn_activity_seed_values.csv`",
"`mechanisms/activity/`；先合并原始能量再计算富集，不能平均已算好的比值"
],
[
"h",
"未受外力的全部 1,056 条通路干预记录",
"`mechanisms/interventions/closed_loop/`；全部十项干预和原四项特异性比较族"
],
[
"i",
"`demo_plus60_contrast_trajectories.csv`、八张帧",
"原始 +60° 转向示例；候选选择审计随数据保存"
],
[
"j",
"`cpg_input_gain_surfaces.npz`",
"`mechanisms/input_gain/`；48 个模型 × 21 × 21 实测网格"
]
]
},
"4": {
"direct_plot_inputs": [
"assets/figure4/demo/MEDIA_PROVENANCE.json",
"assets/figure4/demo/speed_modulation.mp4",
"assets/figure4/simulator_original.png",
"data/figure4/source_data/demo_speed_trace.csv",
"data/figure4/source_data/latest_event_class_ablation_seed.csv",
"data/figure4/source_data/latest_neuron_class_seed_times.csv",
"data/figure4/source_data/latest_parameter_landscapes.npz",
"data/figure4/source_data/main_flatness_mean_grids.npz",
"data/figure4/source_data/scatter_MODEL_MECHANISM_SUMMARY.csv",
"data/figure4/source_data/scatter_performance_associations.csv",
"data/figure4/source_data/short_budget_primary_fixed_budget.csv",
"data/figure4/source_data/short_budget_seed_metrics.csv",
"data/figure4/source_data/v20_acceleration_heatmap.csv",
"data/figure4/source_data/v24_ramp_increment_summary.csv"
],
"panel_map": [
[
"a",
"`assets/figure4/simulator_original.png`"
],
[
"b",
"`short_budget_seed_metrics.csv`；完整三项固定预算终点校正"
],
[
"c",
"`v24_ramp_increment_summary.csv`；原始速度轨迹、固定基线和 0.1 s 窗口"
],
[
"d",
"`latest_neuron_class_seed_times.csv`；`neurons/` 的模型时域灵敏度"
],
[
"e",
"`v20_acceleration_heatmap.csv`；`signed_replays/` 的有符号响应"
],
[
"f",
"`latest_event_class_ablation_seed.csv`；`events/` 中完整目标保持／通道干预响应"
],
[
"g",
"`scatter_MODEL_MECHANISM_SUMMARY.csv`、`scatter_performance_associations.csv`；`probes/`、行为记录和原始模型投影；保留全部 45 项关联"
],
[
"h",
"`demo_speed_trace.csv`、原生视频及逐帧选择信息；显示平滑规则与来源保留"
],
[
"i",
"`latest_parameter_landscapes.npz`；`probes/` 中 40 个模型的 21 × 21 参数扰动面"
]
]
}
}