How far should an embodied system change after unexpected experience? Research OS, Mismatch Lab, and Builder Lab by Seonhye Gu.
Generated:
python scripts/plot_paper1_figures.py
Isaac captures:bash scripts/capture_study1_viewport.sh(GPU) →plot_paper1_from_ee_traces.py
| Fig | File | Tier | Description |
|---|---|---|---|
| 1 | fig1_architecture_overview.png |
A | System stack + same-state CF evaluation pipeline |
| 2 | fig2_same_state_fork.png |
C | Branching at mismatch onset S |
| 3 | fig3_profile_d0.png |
C | D0 five-mode recoverability profile (n=20) |
| 4 | fig4_baseline_overlay.png |
C | Menu vs B2 vs B3 @ fixed S |
| 5 | fig5_occlusion_contrast.png |
C | D0 occluded vs D1 no-occlusion (RQ-P) |
| Table | File | CSV |
|---|---|---|
| 1 D0 profiles | table1_d0_results.png |
tables/table1_d0.csv |
| 2 Program summary | table2_proper_program.png |
tables/table2_program.csv |
| Panel | File | Source |
|---|---|---|
| CONTINUE vs REPLAN schematic | sim_panel_continue_replan.png |
CPU schematic (protocol geometry) |
| Multi-mode menu | sim_panel_multimode.png |
CPU schematic |
| Isaac EE traces | sim_panel_isaac_traces.png |
GPU capture → isaac_captures/ |
Captions: ../fig_captions.md