Add files using upload-large-folder tool

.pytest_cache/.gitignore

@@ -0,0 +1,2 @@
+# Created by pytest automatically.
+*

.pytest_cache/CACHEDIR.TAG

@@ -0,0 +1,4 @@
+Signature: 8a477f597d28d172789f06886806bc55
+# This file is a cache directory tag created by pytest.
+# For information about cache directory tags, see:
+#	https://bford.info/cachedir/spec.html

.pytest_cache/README.md

@@ -0,0 +1,8 @@
+# pytest cache directory #
+
+This directory contains data from the pytest's cache plugin,
+which provides the `--lf` and `--ff` options, as well as the `cache` fixture.
+
+**Do not** commit this to version control.
+
+See [the docs](https://docs.pytest.org/en/stable/how-to/cache.html) for more information.

.pytest_cache/v/cache/lastfailed

@@ -0,0 +1,54 @@
+{
+  "tests/test_encoder/test_geometric.py::test_compact_oriented_box_uses_accumulated_instance_points": true,
+  "tests/test_encoder/test_geometric.py::test_compact_floor_boundaries_preserve_disconnected_regions": true,
+  "tests/test_encoder/test_geometric.py::test_compact_spatial_code_suppresses_co_visible_duplicate_tracks": true,
+  "tests/test_encoder/test_geometric.py::test_compact_oriented_box_combines_observation_extents_by_consensus": true,
+  "tests/test_encoder/test_geometric.py::test_compact_instances_keep_peak_co_visible_hypotheses_by_evidence": true,
+  "tests/test_encoder/test_geometric.py::test_compact_oriented_box_rejects_one_inconsistent_observation": true,
+  "tests/test_encoder/test_geometric.py::test_spatial_code_matches_reference_schema": true,
+  "tests/test_encoder/test_geometric.py::test_compact_spatial_code_exposes_only_reusable_primitives": true,
+  "tests/test_encoder/test_ground_truth.py::test_build_compact_ground_truth_spatial_code_uses_ranks_and_nulls_unranked": true,
+  "tests/test_encoder/test_ground_truth.py::test_build_explicit_ground_truth_spatial_code_is_derived_from_compact": true,
+  "tests/test_B/test_B.py::test_spatial_code_formats_match_the_two_on_disk_schemas": true,
+  "tests/test_B/test_init.py::test_spatial_code_formats_match_the_two_on_disk_schemas": true,
+  "tests/test_B/test_init.py::test_results_dir_defaults_under_workspace_results": true,
+  "tests/test_B/test_prompts.py::test_na_question_prompt_embeds_the_spatial_code_as_text_and_a_post_prompt": true,
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[object_abs_distance]": true,
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[object_counting]": true,
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[object_size_estimation]": true,
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[room_size_estimation]": true,
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_easy]": true,
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_medium]": true,
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_hard]": true,
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_distance]": true,
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[route_planning]": true,
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[obj_appearance_order]": true,
+  "tests/test_B/test_prompts.py::test_default_arguments_reproduce_the_standard_prompt_byte_for_byte": true,
+  "tests/test_B/test_prompts.py::test_yaml_serialization_renders_the_identical_dict": true,
+  "tests/test_B/test_prompts.py::test_yaml_arm_context_line_does_not_claim_json": true,
+  "tests/test_B/test_prompts.py::test_paraphrase_context_line_swaps_only_the_first_line": true,
+  "tests/test_B/test_prompts.py::test_unknown_serialization_rejected": true,
+  "tests/test_B/test_run.py::test_build_record_preserves_every_field_untruncated": true,
+  "tests/test_B/test_run.py::test_run_strip_schema_legend_removes_only_the_legend": true,
+  "tests/test_C/test_init.py::test_results_dir_defaults_under_workspace_results": true,
+  "tests/test_C/test_launch.py::test_launch_routes_overlay_to_overlay_result_tree": true,
+  "tests/test_C/test_prompts.py::test_pre_prompt_mentions_both_frames_and_spatial_code": true,
+  "tests/test_C/test_prompts.py::test_na_question_prompt_layout_is_context_then_code_then_question_then_post_prompt": true,
+  "tests/test_C/test_prompts.py::test_mca_question_prompt_includes_options_and_post_prompt": true,
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[object_abs_distance]": true,
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[object_counting]": true,
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[object_size_estimation]": true,
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[room_size_estimation]": true,
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_easy]": true,
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_medium]": true,
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_hard]": true,
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_distance]": true,
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[route_planning]": true,
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[obj_appearance_order]": true,
+  "tests/test_C/test_run.py::test_build_record_carries_both_frame_and_spatial_code_provenance": true,
+  "tests/test_C/test_sweep.py::test_sweep_forwards_overlay_to_each_launch": true,
+  "tests/test_E/test_sweep.py::test_sweep_defaults_to_base_protocol": true,
+  "tests/test_A/test_init.py::test_results_dir_defaults_under_workspace_results": true,
+  "tests/test_A/test_prompts.py::test_na_question_prompt_matches_vsibench_protocol": true,
+  "tests/test_A/test_prompts.py::test_mca_question_prompt_matches_vsibench_protocol": true
+}

.pytest_cache/v/cache/nodeids

@@ -0,0 +1,334 @@
+[
+  "tests/test_A/test_A.py::test_base_mode_keeps_budgets_unset_without_budget_flags",
+  "tests/test_A/test_A.py::test_base_mode_rejects_thinking_budget_flags[force_budget]",
+  "tests/test_A/test_A.py::test_base_mode_rejects_thinking_budget_flags[reasoning_budget]",
+  "tests/test_A/test_A.py::test_default_frame_selection_is_a_valid_selection",
+  "tests/test_A/test_A.py::test_frame_selections_match_inference_vocabulary",
+  "tests/test_A/test_A.py::test_generation_protocol_matches_vsibench_yaml",
+  "tests/test_A/test_A.py::test_model_paths_cover_every_registered_model",
+  "tests/test_A/test_A.py::test_results_dir_defaults_under_root_results",
+  "tests/test_A/test_A.py::test_thinking_generation_protocol",
+  "tests/test_A/test_A.py::test_thinking_mode_resolves_default_budgets",
+  "tests/test_A/test_frames.py::test_sample_frames_derives_indices_from_real_fps",
+  "tests/test_A/test_frames.py::test_sample_frames_rejects_missing_video",
+  "tests/test_A/test_frames.py::test_sample_frames_rejects_nonpositive_frame_count",
+  "tests/test_A/test_frames.py::test_sample_frames_rejects_unknown_selection",
+  "tests/test_A/test_frames.py::test_sample_frames_returns_pil_images_in_order",
+  "tests/test_A/test_init.py::test_default_frame_selection_is_a_valid_selection",
+  "tests/test_A/test_init.py::test_every_known_question_type_has_one_policy_group",
+  "tests/test_A/test_init.py::test_frame_selections_match_inference_vocabulary",
+  "tests/test_A/test_init.py::test_generation_protocol_matches_vsibench_yaml",
+  "tests/test_A/test_init.py::test_model_paths_cover_every_registered_model",
+  "tests/test_A/test_init.py::test_question_protocol_policy_is_hardcoded_by_group",
+  "tests/test_A/test_init.py::test_results_dir_defaults_under_workspace_results",
+  "tests/test_A/test_init.py::test_thinking_generation_protocol",
+  "tests/test_A/test_launch.py::test_launch_rebuild_forces_pending_even_when_answered",
+  "tests/test_A/test_launch.py::test_launch_rejects_scene_with_no_questions",
+  "tests/test_A/test_launch.py::test_launch_skips_scene_already_fully_answered",
+  "tests/test_A/test_launch.py::test_launcher_imports",
+  "tests/test_A/test_launch.py::test_scenes_dedups_and_preserves_order",
+  "tests/test_A/test_models.py::test_adapter_answer_before_load_model_raises",
+  "tests/test_A/test_models.py::test_adapter_answer_extended_before_load_model_raises",
+  "tests/test_A/test_models.py::test_all_three_models_are_registered",
+  "tests/test_A/test_models.py::test_every_adapter_implements_answer_extended",
+  "tests/test_A/test_models.py::test_get_adapter_binds_the_correct_checkpoint_path",
+  "tests/test_A/test_models.py::test_get_adapter_rejects_unknown_model",
+  "tests/test_A/test_models.py::test_get_adapter_returns_the_right_class_per_model",
+  "tests/test_A/test_models.py::test_internvl_adapter_disables_per_frame_tiling",
+  "tests/test_A/test_models.py::test_native_video_content_uses_one_video_item",
+  "tests/test_A/test_models.py::test_numbered_content_handles_zero_frames",
+  "tests/test_A/test_models.py::test_numbered_content_labels_every_frame_in_order",
+  "tests/test_A/test_models.py::test_unload_clears_model_and_processor",
+  "tests/test_A/test_prompts.py::test_every_mca_question_type_builds_with_options[obj_appearance_order]",
+  "tests/test_A/test_prompts.py::test_every_mca_question_type_builds_with_options[object_rel_direction_easy]",
+  "tests/test_A/test_prompts.py::test_every_mca_question_type_builds_with_options[object_rel_direction_hard]",
+  "tests/test_A/test_prompts.py::test_every_mca_question_type_builds_with_options[object_rel_direction_medium]",
+  "tests/test_A/test_prompts.py::test_every_mca_question_type_builds_with_options[object_rel_distance]",
+  "tests/test_A/test_prompts.py::test_every_mca_question_type_builds_with_options[route_planning]",
+  "tests/test_A/test_prompts.py::test_every_na_question_type_builds_without_options[object_abs_distance]",
+  "tests/test_A/test_prompts.py::test_every_na_question_type_builds_without_options[object_counting]",
+  "tests/test_A/test_prompts.py::test_every_na_question_type_builds_without_options[object_size_estimation]",
+  "tests/test_A/test_prompts.py::test_every_na_question_type_builds_without_options[room_size_estimation]",
+  "tests/test_A/test_prompts.py::test_mca_question_prompt_matches_vsibench_protocol",
+  "tests/test_A/test_prompts.py::test_mca_question_requires_options",
+  "tests/test_A/test_prompts.py::test_na_question_prompt_matches_vsibench_protocol",
+  "tests/test_A/test_prompts.py::test_unknown_question_type_rejected",
+  "tests/test_A/test_prompts.py::test_video_prompt_names_native_video",
+  "tests/test_A/test_run.py::test_build_record_carries_reasoning_fields_when_extended",
+  "tests/test_A/test_run.py::test_build_record_defaults_reasoning_fields_when_not_extended",
+  "tests/test_A/test_run.py::test_build_record_preserves_every_field_untruncated",
+  "tests/test_A/test_run.py::test_load_questions_filters_by_scene",
+  "tests/test_A/test_run.py::test_load_questions_reads_every_row",
+  "tests/test_A/test_run.py::test_load_questions_respects_limit",
+  "tests/test_A/test_run.py::test_results_dir_for_honors_explicit_override",
+  "tests/test_A/test_run.py::test_results_dir_for_isolates_the_two_protocols",
+  "tests/test_A/test_run.py::test_results_dir_for_keeps_protocols_together",
+  "tests/test_A/test_run.py::test_results_dir_for_matches_established_dimension_nesting",
+  "tests/test_A/test_run.py::test_scalar_score_rejects_unknown_question_type",
+  "tests/test_A/test_run.py::test_scalar_score_returns_metric_name_and_value",
+  "tests/test_A/test_run.py::test_video_record_has_no_frame_count_in_condition",
+  "tests/test_A/test_run.py::test_video_results_use_video_branch",
+  "tests/test_A/test_run.py::test_write_question_result_writes_one_json_file_per_question",
+  "tests/test_A/test_sweep.py::test_build_plan_covers_every_combination",
+  "tests/test_A/test_sweep.py::test_build_plan_orders_by_frame_count_first",
+  "tests/test_A/test_sweep.py::test_build_plan_with_all_registered_models",
+  "tests/test_A/test_sweep.py::test_parse_csv_choice_expands_all",
+  "tests/test_A/test_sweep.py::test_parse_csv_choice_rejects_empty",
+  "tests/test_A/test_sweep.py::test_parse_csv_choice_rejects_unknown_value",
+  "tests/test_A/test_sweep.py::test_parse_csv_choice_splits_and_dedups",
+  "tests/test_A/test_sweep.py::test_parse_frame_counts_rejects_non_integer",
+  "tests/test_A/test_sweep.py::test_parse_frame_counts_rejects_nonpositive",
+  "tests/test_A/test_sweep.py::test_parse_frame_counts_splits_and_dedups",
+  "tests/test_B/test_B.py::test_depth_and_tracking_reuse_encoder_config_vocabulary",
+  "tests/test_B/test_B.py::test_input_selections_match_harness_a_vocabulary",
+  "tests/test_B/test_B.py::test_results_dir_defaults_under_root_results",
+  "tests/test_B/test_B.py::test_reuses_harness_a_model_paths_and_generation_protocol",
+  "tests/test_B/test_B.py::test_spatial_code_formats_is_explicit_only",
+  "tests/test_B/test_B.py::test_spatial_code_formats_match_the_two_on_disk_schemas",
+  "tests/test_B/test_init.py::test_depth_and_tracking_reuse_encoder_config_vocabulary",
+  "tests/test_B/test_init.py::test_input_selections_match_harness_a_vocabulary",
+  "tests/test_B/test_init.py::test_results_dir_defaults_under_workspace_results",
+  "tests/test_B/test_init.py::test_reuses_harness_a_model_paths_and_generation_protocol",
+  "tests/test_B/test_init.py::test_spatial_code_formats_is_explicit_only",
+  "tests/test_B/test_init.py::test_spatial_code_formats_match_the_two_on_disk_schemas",
+  "tests/test_B/test_launch.py::test_launch_rebuild_forces_pending_even_when_answered",
+  "tests/test_B/test_launch.py::test_launch_rejects_question_id_filter_that_matches_nothing",
+  "tests/test_B/test_launch.py::test_launch_skips_scene_already_fully_answered",
+  "tests/test_B/test_launch.py::test_launcher_imports",
+  "tests/test_B/test_prompts.py::test_default_arguments_reproduce_the_standard_prompt_byte_for_byte",
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[obj_appearance_order]",
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_easy]",
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_hard]",
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_medium]",
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[object_rel_distance]",
+  "tests/test_B/test_prompts.py::test_every_mca_question_type_builds[route_planning]",
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[object_abs_distance]",
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[object_counting]",
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[object_size_estimation]",
+  "tests/test_B/test_prompts.py::test_every_na_question_type_builds[room_size_estimation]",
+  "tests/test_B/test_prompts.py::test_mca_question_prompt_includes_options_and_matches_harness_a_post_prompt",
+  "tests/test_B/test_prompts.py::test_mca_question_requires_options",
+  "tests/test_B/test_prompts.py::test_na_question_prompt_embeds_the_spatial_code_as_text_and_a_post_prompt",
+  "tests/test_B/test_prompts.py::test_no_frames_language_in_pre_prompt",
+  "tests/test_B/test_prompts.py::test_paraphrase_context_line_swaps_only_the_first_line",
+  "tests/test_B/test_prompts.py::test_unknown_question_type_rejected",
+  "tests/test_B/test_prompts.py::test_unknown_serialization_rejected",
+  "tests/test_B/test_prompts.py::test_yaml_arm_context_line_does_not_claim_json",
+  "tests/test_B/test_prompts.py::test_yaml_serialization_renders_the_identical_dict",
+  "tests/test_B/test_run.py::test_build_record_carries_reasoning_fields_when_forced",
+  "tests/test_B/test_run.py::test_build_record_preserves_every_field_untruncated",
+  "tests/test_B/test_run.py::test_results_dir_for_honors_explicit_override",
+  "tests/test_B/test_run.py::test_results_dir_for_isolates_the_two_protocols",
+  "tests/test_B/test_run.py::test_results_dir_for_keeps_protocols_together",
+  "tests/test_B/test_run.py::test_results_dir_for_matches_established_dimension_nesting",
+  "tests/test_B/test_run.py::test_run_strip_schema_legend_removes_only_the_legend",
+  "tests/test_B/test_run.py::test_video_record_has_no_frame_count_in_condition",
+  "tests/test_B/test_run.py::test_video_results_use_video_branch",
+  "tests/test_B/test_run.py::test_write_question_result_writes_one_json_file_per_question",
+  "tests/test_B/test_spatial_codes.py::test_load_spatial_code_raises_clearly_when_missing",
+  "tests/test_B/test_spatial_codes.py::test_load_spatial_code_rejects_unknown_format",
+  "tests/test_B/test_spatial_codes.py::test_load_spatial_code_returns_dict_and_path",
+  "tests/test_B/test_spatial_codes.py::test_spatial_code_path_uses_tracking_frames_hierarchy",
+  "tests/test_B/test_sweep.py::test_build_plan_covers_every_combination",
+  "tests/test_B/test_sweep.py::test_build_plan_orders_by_frame_count_first",
+  "tests/test_B/test_sweep.py::test_build_plan_sweeps_depth_and_tracking_too",
+  "tests/test_B/test_sweep.py::test_build_plan_with_all_registered_models",
+  "tests/test_B/test_sweep.py::test_sweep_parser_rejects_unknown_depth",
+  "tests/test_B/test_sweep.py::test_sweep_parser_rejects_unknown_spatial_code_format",
+  "tests/test_B/test_sweep.py::test_sweep_parser_rejects_unknown_tracking",
+  "tests/test_C/test_C.py::test_depth_and_tracking_reuse_encoder_config_vocabulary",
+  "tests/test_C/test_C.py::test_input_selections_are_one_shared_vocabulary_with_a_and_b",
+  "tests/test_C/test_C.py::test_results_dir_defaults_under_root_results",
+  "tests/test_C/test_C.py::test_reuses_harness_a_model_paths_and_generation_protocol",
+  "tests/test_C/test_init.py::test_depth_and_tracking_reuse_encoder_config_vocabulary",
+  "tests/test_C/test_init.py::test_input_selections_are_one_shared_vocabulary_with_a_and_b",
+  "tests/test_C/test_init.py::test_results_dir_defaults_under_workspace_results",
+  "tests/test_C/test_init.py::test_reuses_harness_a_model_paths_and_generation_protocol",
+  "tests/test_C/test_launch.py::test_launch_rebuild_forces_pending_even_when_answered",
+  "tests/test_C/test_launch.py::test_launch_routes_overlay_to_overlay_result_tree",
+  "tests/test_C/test_launch.py::test_launch_skips_scene_already_fully_answered",
+  "tests/test_C/test_launch.py::test_launcher_imports",
+  "tests/test_C/test_overlay.py::test_instance_ids_adds_stable_one_based_labels_without_mutating_input",
+  "tests/test_C/test_overlay.py::test_label_positions_parse_meter_strings_in_code_order",
+  "tests/test_C/test_overlay.py::test_load_cached_frames_requires_complete_png_set_and_labels",
+  "tests/test_C/test_overlay.py::test_load_or_create_overlay_code_reuses_existing_file",
+  "tests/test_C/test_overlay.py::test_load_or_create_overlay_code_saves_missing_file",
+  "tests/test_C/test_overlay.py::test_overlay_spatial_code_path_lives_under_overlay_root",
+  "tests/test_C/test_overlay.py::test_save_cached_frames_writes_pngs_and_labels",
+  "tests/test_C/test_overlay.py::test_stamp_frames_serves_complete_cache_without_loading_dependencies",
+  "tests/test_C/test_overlay.py::test_stamp_frames_uses_raw_sam3_boxes_and_does_not_mutate_inputs",
+  "tests/test_C/test_overlay_launch.py::test_available_cpu_count_honors_positive_environment",
+  "tests/test_C/test_overlay_launch.py::test_has_dependencies_requires_spatial_code_and_sam3_cache",
+  "tests/test_C/test_overlay_launch.py::test_has_dependencies_treats_missing_code_as_ineligible",
+  "tests/test_C/test_overlay_launch.py::test_launch_does_not_skip_frames_cache_when_overlay_code_is_missing",
+  "tests/test_C/test_overlay_launch.py::test_launch_skips_missing_and_already_cached_without_pool",
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[obj_appearance_order]",
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_easy]",
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_hard]",
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_direction_medium]",
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[object_rel_distance]",
+  "tests/test_C/test_prompts.py::test_every_mca_question_type_builds[route_planning]",
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[object_abs_distance]",
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[object_counting]",
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[object_size_estimation]",
+  "tests/test_C/test_prompts.py::test_every_na_question_type_builds[room_size_estimation]",
+  "tests/test_C/test_prompts.py::test_mca_question_prompt_includes_options_and_post_prompt",
+  "tests/test_C/test_prompts.py::test_mca_question_requires_options",
+  "tests/test_C/test_prompts.py::test_na_question_prompt_layout_is_context_then_code_then_question_then_post_prompt",
+  "tests/test_C/test_prompts.py::test_pre_prompt_mentions_both_frames_and_spatial_code",
+  "tests/test_C/test_prompts.py::test_unknown_question_type_rejected",
+  "tests/test_C/test_prompts.py::test_video_prompt_names_native_video",
+  "tests/test_C/test_run.py::test_build_record_carries_both_frame_and_spatial_code_provenance",
+  "tests/test_C/test_run.py::test_build_record_carries_reasoning_fields_when_forced",
+  "tests/test_C/test_run.py::test_results_dir_for_honors_explicit_override",
+  "tests/test_C/test_run.py::test_results_dir_for_matches_established_dimension_nesting",
+  "tests/test_C/test_run.py::test_results_dir_for_places_overlay_under_C_results",
+  "tests/test_C/test_run.py::test_video_record_has_no_frame_count_in_condition",
+  "tests/test_C/test_run.py::test_video_results_use_video_branch",
+  "tests/test_C/test_run.py::test_write_question_result_routes_overlay_default_to_overlay_tree",
+  "tests/test_C/test_run.py::test_write_question_result_writes_one_json_file_per_question",
+  "tests/test_C/test_sweep.py::test_build_plan_covers_every_combination",
+  "tests/test_C/test_sweep.py::test_build_plan_orders_by_frame_count_first",
+  "tests/test_C/test_sweep.py::test_build_plan_sweeps_depth_and_tracking_too",
+  "tests/test_C/test_sweep.py::test_build_plan_with_all_registered_models",
+  "tests/test_C/test_sweep.py::test_sweep_forwards_overlay_to_each_launch",
+  "tests/test_C/test_sweep.py::test_sweep_parser_rejects_unknown_depth",
+  "tests/test_C/test_sweep.py::test_sweep_parser_rejects_unknown_spatial_code_format",
+  "tests/test_C/test_sweep.py::test_sweep_parser_rejects_unknown_tracking",
+  "tests/test_E/test_launch.py::test_launch_protocols_share_result_root",
+  "tests/test_E/test_launch.py::test_launch_protocols_use_separate_result_roots",
+  "tests/test_E/test_launch.py::test_launch_rebuild_forces_pending_even_when_answered",
+  "tests/test_E/test_launch.py::test_launch_skips_scene_already_fully_answered",
+  "tests/test_E/test_launch.py::test_launcher_imports",
+  "tests/test_E/test_run.py::test_build_record_has_no_scene_input_provenance",
+  "tests/test_E/test_run.py::test_results_dir_for_honors_explicit_override",
+  "tests/test_E/test_run.py::test_results_dir_for_keeps_protocols_together",
+  "tests/test_E/test_run.py::test_results_dir_for_matches_model_only",
+  "tests/test_E/test_run.py::test_write_question_result_writes_one_json_file_per_question",
+  "tests/test_E/test_sweep.py::test_sweep_defaults_to_base_protocol",
+  "tests/test_E/test_sweep.py::test_sweep_imports",
+  "tests/test_E/test_sweep.py::test_sweep_parser_rejects_unknown_model",
+  "tests/test_E/test_sweep.py::test_sweep_runs_every_model_through_launch",
+  "tests/test_E/test_sweep.py::test_sweep_uses_the_shared_question_policy_without_a_global_protocol",
+  "tests/test_F/test_F.py::test_results_dir_can_be_overridden_by_environment",
+  "tests/test_F/test_F.py::test_sources_and_default_are_declared",
+  "tests/test_F/test_launch.py::test_launch_entrypoint_exposes_run_main",
+  "tests/test_F/test_run.py::test_perceived_layout_contains_every_input_axis",
+  "tests/test_F/test_run.py::test_results_default_under_root",
+  "tests/test_F/test_run.py::test_video_results_use_video_branch",
+  "tests/test_F/test_sweep.py::test_csv_expands_all_dedups_and_rejects_unknowns",
+  "tests/test_analysis/test_A_reports.py::TestAReports::test_A_report_and_controlled_frame_comparison",
+  "tests/test_analysis/test_B_reports.py::TestBReports::test_B_report_contains_spatial_cell",
+  "tests/test_analysis/test_C_reports.py::TestCReports::test_C_report_is_exported",
+  "tests/test_analysis/test_F_reports.py::TestFReports::test_F_legacy_and_future_perceived_layouts",
+  "tests/test_analysis/test_analysis.py::TestAnalysisDirectory::test_arbitrary_subset_exports_letter_and_combined_files",
+  "tests/test_analysis/test_letters_reports.py::TestLettersReports::test_arbitrary_subset_and_combined_name",
+  "tests/test_analysis/test_letters_reports.py::TestLettersReports::test_folded_statistical_and_solver_helpers",
+  "tests/test_analysis/test_letters_reports.py::TestLettersReports::test_manifest_generated_at_is_deterministic_by_default",
+  "tests/test_analysis/test_letters_reports.py::TestLettersReports::test_pair_restriction",
+  "tests/test_backup.py::test_backup_dry_run_skips_empty_targets_and_never_imports_hub",
+  "tests/test_backup.py::test_local_paths_keep_results_under_results_host",
+  "tests/test_backup.py::test_resolve_targets_expands_all_without_uploading_unknowns",
+  "tests/test_backup.py::test_target_root_rejects_mixed_workspace_and_results_roots",
+  "tests/test_encoder/test_adapters.py::test_adapt_sam3_depth_anything_3_decodes_masks_and_backprojects",
+  "tests/test_encoder/test_adapters.py::test_adapt_segvggt_reads_flat_npz",
+  "tests/test_encoder/test_adapters.py::test_adapt_segvggt_rejects_missing_npz_fields",
+  "tests/test_encoder/test_adapters.py::test_adapt_segvggt_requires_existing_cache",
+  "tests/test_encoder/test_adapters.py::test_adapter_owned_raw_cache_locations",
+  "tests/test_encoder/test_adapters.py::test_adapters_default_to_root_data_caches",
+  "tests/test_encoder/test_adapters.py::test_backproject_resizes_sam3_mask_to_da3_depth_shape",
+  "tests/test_encoder/test_adapters.py::test_fusion_adapter_resolves_two_native_model_directories",
+  "tests/test_encoder/test_adapters.py::test_native_fusion_times_are_measured_in_seconds",
+  "tests/test_encoder/test_adapters.py::test_native_sam3_decodes_lossless_tracking_cache",
+  "tests/test_encoder/test_adapters.py::test_native_sam3_decodes_prompt_keyed_independent_frames",
+  "tests/test_encoder/test_adapters.py::test_native_sam3_preserves_tracked_object_ids",
+  "tests/test_encoder/test_adapters.py::test_registry_decodes_native_segvggt_dictionary",
+  "tests/test_encoder/test_adapters.py::test_spatial_code_format_validation",
+  "tests/test_encoder/test_adapters.py::test_validate_identifies_empty_scene",
+  "tests/test_encoder/test_adapters.py::test_validate_normalizes_canonical_geometry",
+  "tests/test_encoder/test_adapters.py::test_validate_rejects_invalid_geometry[scene0-TypeError]",
+  "tests/test_encoder/test_adapters.py::test_validate_rejects_invalid_geometry[scene1-ValueError]",
+  "tests/test_encoder/test_adapters.py::test_validate_rejects_invalid_geometry[scene2-ValueError]",
+  "tests/test_encoder/test_config.py::test_encoder_paths_mirror_all_dimensions",
+  "tests/test_encoder/test_config.py::test_encoder_paths_reject_unknown_spatial_code_format",
+  "tests/test_encoder/test_config.py::test_ground_truth_paths_have_no_depth_tracking_input_frames_axis",
+  "tests/test_encoder/test_config.py::test_ground_truth_paths_reject_unknown_spatial_code_format",
+  "tests/test_encoder/test_encoder.py::test_backproject_frame_applies_intrinsics_pose_and_confidence",
+  "tests/test_encoder/test_encoder.py::test_backproject_frame_returns_typed_empty_array",
+  "tests/test_encoder/test_encoder.py::test_closest_distance_uses_point_cloud_distance",
+  "tests/test_encoder/test_encoder.py::test_depth_edges_handles_small_and_discontinuous_frames",
+  "tests/test_encoder/test_encoder.py::test_relative_direction_modes",
+  "tests/test_encoder/test_encoder.py::test_robust_centroid_extent_returns_sorted_dimensions",
+  "tests/test_encoder/test_geometric.py::test_compact_floor_boundaries_preserve_disconnected_regions",
+  "tests/test_encoder/test_geometric.py::test_compact_instances_keep_peak_co_visible_hypotheses_by_evidence",
+  "tests/test_encoder/test_geometric.py::test_compact_oriented_box_combines_observation_extents_by_consensus",
+  "tests/test_encoder/test_geometric.py::test_compact_oriented_box_rejects_one_inconsistent_observation",
+  "tests/test_encoder/test_geometric.py::test_compact_oriented_box_uses_accumulated_instance_points",
+  "tests/test_encoder/test_geometric.py::test_compact_spatial_code_exposes_only_reusable_primitives",
+  "tests/test_encoder/test_geometric.py::test_compact_spatial_code_merges_revisit_instances_and_keeps_earliest_time",
+  "tests/test_encoder/test_geometric.py::test_compact_spatial_code_suppresses_co_visible_duplicate_tracks",
+  "tests/test_encoder/test_geometric.py::test_dump_spatial_code",
+  "tests/test_encoder/test_geometric.py::test_dumped_json_preserves_schema",
+  "tests/test_encoder/test_geometric.py::test_exact_math_is_integrated_into_geometric_module",
+  "tests/test_encoder/test_geometric.py::test_explicit_is_a_derivation_of_compact",
+  "tests/test_encoder/test_geometric.py::test_explicit_spatial_code_remains_the_default",
+  "tests/test_encoder/test_geometric.py::test_floor_level_v1_v2_math_is_shared",
+  "tests/test_encoder/test_geometric.py::test_position_reader_accepts_current_and_legacy_formatting",
+  "tests/test_encoder/test_geometric.py::test_raw_bundle_dispatches_to_explicit_derivation",
+  "tests/test_encoder/test_geometric.py::test_spatial_code_matches_reference_schema",
+  "tests/test_encoder/test_ground_truth.py::test_appearance_order_ranks_topological_across_multiple_questions",
+  "tests/test_encoder/test_ground_truth.py::test_build_all_respects_explicit_scene_list_and_formats",
+  "tests/test_encoder/test_ground_truth.py::test_build_all_writes_every_scene_and_format",
+  "tests/test_encoder/test_ground_truth.py::test_build_and_write_writes_to_the_ground_truth_path",
+  "tests/test_encoder/test_ground_truth.py::test_build_compact_ground_truth_spatial_code_nulls_untimed_classes",
+  "tests/test_encoder/test_ground_truth.py::test_build_compact_ground_truth_spatial_code_raises_for_unknown_scene",
+  "tests/test_encoder/test_ground_truth.py::test_build_compact_ground_truth_spatial_code_uses_ranks_and_nulls_unranked",
+  "tests/test_encoder/test_ground_truth.py::test_build_explicit_ground_truth_spatial_code_is_derived_from_compact",
+  "tests/test_encoder/test_ground_truth.py::test_build_ground_truth_spatial_code_dispatches_by_format",
+  "tests/test_encoder/test_ground_truth.py::test_build_ground_truth_spatial_code_rejects_unknown_format",
+  "tests/test_encoder/test_ground_truth.py::test_floor_level_accounts_for_tilted_box_support",
+  "tests/test_encoder/test_ground_truth.py::test_floor_level_defaults_to_zero_for_empty_scene",
+  "tests/test_encoder/test_ground_truth.py::test_floor_level_is_the_lowest_box_support_across_every_instance",
+  "tests/test_encoder/test_ground_truth.py::test_gt_floor_boundary_polygon_area_matches_room_size",
+  "tests/test_encoder/test_ground_truth.py::test_gt_oriented_box_rebases_height_above_floor_and_passes_xy_through",
+  "tests/test_encoder/test_ground_truth.py::test_gt_oriented_box_renormalizes_orientation_vectors",
+  "tests/test_encoder/test_ground_truth.py::test_load_meta_info_merges_all_three_datasets_and_tags_dataset",
+  "tests/test_encoder/test_ground_truth.py::test_scenes_returns_sorted_meta_info_keys",
+  "tests/test_encoder/test_init.py::test_encoder_package_imports_without_data_or_checkpoints",
+  "tests/test_encoder/test_launch.py::test_encoder_launcher_imports",
+  "tests/test_encoder/test_render.py::test_render_exposes_tracking_and_format_dimensions",
+  "tests/test_encoder/test_run.py::test_cache_or_load_exposes_explicit_dimensions",
+  "tests/test_encoder/test_run.py::test_video_mode_preserves_depth_and_tracking",
+  "tests/test_encoder/test_run.py::test_video_mode_selects_caches_for_requested_axes",
+  "tests/test_harness/test_harness.py::test_harness_package_imports_without_side_effects",
+  "tests/test_symbolic/test_adapters.py::test_adapter_rejects_format_mismatch",
+  "tests/test_symbolic/test_adapters.py::test_compact_adapter_derives_solver_values_from_primitives",
+  "tests/test_symbolic/test_adapters.py::test_compact_adapter_ignores_none_instances_within_a_partially_timed_class",
+  "tests/test_symbolic/test_adapters.py::test_compact_adapter_treats_none_first_visible_time_as_unknown",
+  "tests/test_symbolic/test_adapters.py::test_explicit_adapter_preserves_existing_solver_shape",
+  "tests/test_symbolic/test_adapters.py::test_floor_area_sums_disconnected_polygons_and_subtracts_all_holes",
+  "tests/test_symbolic/test_adapters.py::test_oriented_box_distance_is_surface_to_surface",
+  "tests/test_symbolic/test_adapters.py::test_oriented_box_distance_uses_corresponding_rotated_axes",
+  "tests/test_symbolic/test_adapters.py::test_polygon_area_implicitly_closes_ordered_boundary",
+  "tests/test_symbolic/test_adapters.py::test_symbolic_solver_answers_all_direction_levels_from_compact_code",
+  "tests/test_symbolic/test_adapters.py::test_symbolic_solver_answers_numeric_categories_from_compact_code",
+  "tests/test_symbolic/test_adapters.py::test_symbolic_solver_answers_relative_distance_and_order_from_compact_code",
+  "tests/test_symbolic/test_adapters.py::test_symbolic_solver_answers_route_planning_from_compact_code",
+  "tests/test_symbolic/test_launch.py::test_symbolic_launcher_imports",
+  "tests/test_symbolic/test_run.py::test_fetch_spatial_code_uses_selected_format_adapter",
+  "tests/test_symbolic/test_run.py::test_results_dir_for_selection_uses_ground_truth_layout",
+  "tests/test_symbolic/test_run.py::test_select_ground_truth_spatial_codes_points_at_the_ground_truth_directory",
+  "tests/test_symbolic/test_run.py::test_select_ground_truth_spatial_codes_rejects_unknown_format",
+  "tests/test_symbolic/test_run.py::test_select_spatial_codes_clears_ground_truth_flag",
+  "tests/test_symbolic/test_run.py::test_symbolic_selection_isolates_compact_codes",
+  "tests/test_symbolic/test_run.py::test_symbolic_selection_uses_new_hierarchy",
+  "tests/test_symbolic/test_run.py::test_write_question_result_nulls_perception_fields_under_ground_truth",
+  "tests/test_symbolic/test_solver.py::test_answer_snapshots_operation_counts_per_question",
+  "tests/test_symbolic/test_solver.py::test_direct_numeric_answers",
+  "tests/test_symbolic/test_solver.py::test_direction_answers_use_floor_coordinates",
+  "tests/test_symbolic/test_solver.py::test_dispatch_returns_none_for_unknown_or_missing_data",
+  "tests/test_symbolic/test_solver.py::test_multiple_choice_distance_and_order_answers",
+  "tests/test_symbolic/test_solver.py::test_route_planning_chains_turns",
+  "tests/test_symbolic/test_solver.py::test_unit_parsers_accept_strings_and_numbers",
+  "tests/test_symbolic/test_symbolic.py::test_symbolic_folder_modules_are_wired_together"
+]

README.md

@@ -60,7 +60,7 @@ expected locations:
 | Workspace source | `/workspace` | Python packages, tests, setup, README |
 | VSI-Bench dataset | `/root/data/VSI-Bench` | `test.jsonl` plus videos under dataset folders |
 | Official VSI scorer and meta info | `/root/data/thinking-in-space` | `lmms_eval/tasks/vsibench/utils.py` and `data/meta_info` |
-| SAM3/DA3/raw caches | `/root/data/caches` | Input to encoder and overlay stamping |
+| SAM3/DA3/raw caches | `/root/data/caches` | Input to encoder |
 | Spatial codes | `/root/data/spatial codes` | Set with `VSI_CODES`; tests do not depend on it |
 | Model repos/checkpoints | `/root/models` | SAM3, DA3, VLM checkpoints, optional SegVGGT |
 | Harness results | `/root/results` | One JSON per question |
@@ -87,10 +87,9 @@ Regenerating SAM3/DA3 raw caches themselves is not guaranteed deterministic.
 
 1. Run perception caches if needed with `inference.run` or `inference.launch`.
 2. Build perceived spatial codes with `encoder.run` or `encoder.launch`.
-3. Build ground-truth spatial codes with `encoder.ground_truth`.
-4. Run VLM harnesses A-E or symbolic harness F.
-5. Regenerate reports with `analysis.letters_reports`.
-6. Back up selected outputs with `backup.py`.
+3. Run VLM harnesses A-C or symbolic harness F.
+4. Regenerate reports with `analysis.letters_reports`.
+5. Back up selected outputs with `backup.py`.
 
 ## Command Reference
 
@@ -110,7 +109,6 @@ python backup.py A,B,C --repo owner/dataset
 ```
 
 `backup.py` targets are defined in `TARGETS`: `code`, `reports`, `spatial-codes`,
-`caches`, `A`, `B`, `C`, `D`, `E`, `F`, `calibration`, and `corruption`.
 Dry-run mode does not import or call `huggingface_hub`.
 
 ### Inference: Raw Model Caches
@@ -137,25 +135,25 @@ Important files:
 
 ```bash
 # One scene from existing SAM3/DA3 caches.
-python -m encoder.run SCENE --depth metric --tracking tracking --input uniform --frames 32 --format explicit
-python -m encoder.run SCENE --depth metric --tracking tracking --input uniform --frames 32 --format compact
+python -m encoder.run SCENE --depth metric --tracking tracking --input uniform --frames 32
 
 # Batch over every manifest scene with required caches.
-python -m encoder.launch --depth metric --tracking tracking --input uniform --frames 32 --format explicit
+python -m encoder.launch --depth metric --tracking tracking --input uniform --frames 32
+
+# Full-video mode; depth and tracking remain independently selected.
+python -m encoder.run SCENE --depth relative --tracking tracking --video
+python -m encoder.launch --depth relative --tracking tracking --video
 
-# Ground-truth codes from thinking-in-space annotations, no SAM3/DA3.
-python -m encoder.ground_truth --formats explicit,compact
 ```
 
 Important files:
 
 - `encoder/config.py`: path construction and validated axes
 - `encoder/adapters.py`: raw-cache readers into canonical geometry
-- `encoder/geometric.py`: compact and explicit spatial-code construction math
-- `encoder/render.py`: writes final spatial-code JSON
+- `encoder/geometric.py`: spatial-code construction math
+- `encoder/render.py`: writes final explicit spatial-code JSON
 - `encoder/run.py`: loads/verifies raw caches and builds one scene
 - `encoder/launch.py`: CPU-parallel batch driver
-- `encoder/ground_truth.py`: ground-truth compact/explicit code builder
 
 ### Experiments: Geometry Hypotheses
 
@@ -207,16 +205,21 @@ Important files:
 - `symbolic/run.py`: scores one scene and writes per-question JSON
 - `symbolic/launch.py`: multi-scene symbolic orchestration
 
-### Harness A: Frames Only
+### Harness A: Frames or Native Video
+
+The shared policy is configured in `harness/A/__init__.py` as `QUESTION_PROTOCOLS`. Edit that mapping once to switch either question group. Both groups write to the same configuration folder; each result JSON records its own `question_group` and `protocol`.
 
 ```bash
 python -m harness.A.run --model qwen3.5-4b --frame-selection uniform --frames 32 --scene SCENE
 python -m harness.A.launch --model qwen3.5-4b --frame-selection uniform --frames 32 --scenes scene1,scene2
 python -m harness.A.sweep --models all --frame-selections all --frames 16,32
+python -m harness.A.run --model qwen3.5-4b --video --scene SCENE
+python -m harness.A.launch --model qwen3.5-4b --video
+python -m harness.A.sweep --models all --video
+
 ```
 
-Common flags: `--results-dir`, `--rebuild`, `--limit`, `--device`, `--no-write`,
-`--extended`, `--reasoning-budget`, `--force-budget`.
+Question protocols are hardcoded centrally: numerical questions use `base`; multiple-choice questions use `thinking`. `--reasoning-budget` and `--force-budget` affect only thinking questions. Other flags include `--results-dir`, `--rebuild`, `--limit`, `--device`, and `--no-write`.
 
 Files:
 
@@ -231,112 +234,53 @@ Files:
 ### Harness B: Spatial Code Text Only
 
 ```bash
-python -m harness.B.run --model qwen3.5-4b --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32 --scene SCENE
-python -m harness.B.launch --model qwen3.5-4b --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32
-python -m harness.B.sweep --models all --spatial-code-formats all --depths metric --trackings tracking --input-selections uniform --frames 32
+python -m harness.B.run --model qwen3.5-4b --depth metric --tracking tracking --input-selection uniform --frames 32 --scene SCENE
+python -m harness.B.launch --model qwen3.5-4b --depth metric --tracking tracking --input-selection uniform --frames 32
+python -m harness.B.sweep --models all --depths metric --trackings tracking --input-selections uniform --frames 32
+python -m harness.B.run --model qwen3.5-4b --depth metric --tracking tracking --video --scene SCENE
+python -m harness.B.launch --model qwen3.5-4b --depth metric --tracking tracking --video
+python -m harness.B.sweep --models all --depths metric --trackings tracking --video
 ```
 
-Common flags: `--base-protocol`, `--truncated-budget`, `--reasoning-budget`,
-`--force-budget`, `--results-dir`, `--rebuild`, `--thinking` where exposed.
+Question protocols use the same hardcoded numerical=`base`, multiple-choice=`thinking` policy. Budget flags affect only thinking questions. Other common flags are `--results-dir` and `--rebuild`.
 
 Files:
 
 - `harness/B/__init__.py`: B constants and result root
-- `harness/B/spatial_codes.py`: loads perceived compact/explicit JSON
+- `harness/B/spatial_codes.py`: loads perceived explicit JSON
 - `harness/B/prompts.py`: spatial-code-only prompt construction
 - `harness/B/run.py`: one model/config/scene
 - `harness/B/launch.py`: persistent GPU workers for one config
-- `harness/B/sweep.py`: grid over model/code/depth/tracking/input/frame axes
+- `harness/B/sweep.py`: grid over model/depth/tracking/input/frame axes
 
 ### Harness C: Frames Plus Spatial Code
 
 ```bash
-python -m harness.C.run --model qwen3.5-4b --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32 --scene SCENE
-python -m harness.C.launch --model qwen3.5-4b --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32
-python -m harness.C.sweep --models all --spatial-code-formats all --depths metric --trackings tracking --input-selections uniform --frames 32
-```
-
-Overlay/correspondence variants:
-
-```bash
-# Strong correspondence arm: stamped frames + matching saved overlay code.
-python -m harness.C.overlay_launch --depth metric --tracking tracking --input uniform --frames 32
-python -m harness.C.run --model qwen3.5-4b --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32 --overlay-ids --no-schema-legend --scene SCENE
-python -m harness.C.launch --model qwen3.5-4b --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32 --overlay-ids --no-schema-legend
-
-# Weak correspondence arm: add first-visible frame pointers to code.
-python -m harness.C.run --model qwen3.5-4b --spatial-code-format explicit --frame-linked-code --results-dir /root/results/C/frame-linked --scene SCENE
-```
-
-Overlay code artifacts are saved under:
-
-```text
-$VSI_CODES/overlay/sam3+depth-anything-3/<depth>/<tracking>/<input>/<frames>/explicit/<scene>.json
-```
-
-Overlay frames are cached under:
-
-```text
-$VSI_CACHE_ROOT/overlay-frames/<depth>/<tracking>/<input>/<frames>/<scene>/
+python -m harness.C.run --model qwen3.5-4b --depth metric --tracking tracking --input-selection uniform --frames 32 --scene SCENE
+python -m harness.C.launch --model qwen3.5-4b --depth metric --tracking tracking --input-selection uniform --frames 32
+python -m harness.C.sweep --models all --depths metric --trackings tracking --input-selections uniform --frames 32
+python -m harness.C.run --model qwen3.5-4b --depth metric --tracking tracking --video --scene SCENE
+python -m harness.C.launch --model qwen3.5-4b --depth metric --tracking tracking --video
+python -m harness.C.sweep --models all --depths metric --trackings tracking --video
 ```
 
 Files:
 
 - `harness/C/__init__.py`: C constants and result root
 - `harness/C/prompts.py`: combined frames + code prompt construction
-- `harness/C/overlay.py`: Set-of-Marks frame stamping and overlay-code persistence
-- `harness/C/overlay_launch.py`: pre-generates overlay frame/code artifacts
 - `harness/C/run.py`: one model/config/scene
 - `harness/C/launch.py`: persistent GPU workers for one config
-- `harness/C/sweep.py`: grid over model/code/depth/tracking/input/frame axes
-
-### Harness D: Ground-Truth Spatial Code
-
-```bash
-python -m harness.D.run --model qwen3.5-4b --spatial-code-format explicit --scene SCENE
-python -m harness.D.launch --model qwen3.5-4b --spatial-code-format explicit
-python -m harness.D.sweep --models all --spatial-code-formats all
-
-# Perfect-information symbolic ceiling, no VLM.
-python -m harness.D.symbolic_eval --spatial-code-format explicit
-```
-
-Files:
-
-- `harness/D/__init__.py`: D constants and result root
-- `harness/D/spatial_codes.py`: loads ground-truth compact/explicit JSON
-- `harness/D/prompts.py`: ground-truth-code prompt construction
-- `harness/D/run.py`: one model/config/scene
-- `harness/D/launch.py`: persistent GPU workers for one config
-- `harness/D/sweep.py`: grid over models and formats
-- `harness/D/symbolic_eval.py`: symbolic solver on ground-truth codes
-
-### Harness E: Blind Floor
-
-```bash
-python -m harness.E.run --model qwen3.5-4b --scene SCENE
-python -m harness.E.launch --model qwen3.5-4b
-python -m harness.E.sweep --models all
-```
-
-Common flags: `--extended`, `--reasoning-budget`, `--force-budget`,
-`--results-dir`, `--rebuild`, `--limit`, `--device`, `--no-write`.
-
-Files:
-
-- `harness/E/__init__.py`: E constants and result root
-- `harness/E/prompts.py`: question-only prompt construction
-- `harness/E/run.py`: one model/config/scene
-- `harness/E/launch.py`: persistent GPU workers for one config
-- `harness/E/sweep.py`: grid over models/protocols
+- `harness/C/sweep.py`: grid over model/depth/tracking/input/frame axes
 
 ### Harness F: Symbolic Solver As A Harness
 
 ```bash
-python -m harness.F.run --source perceived --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32 --scene SCENE
-python -m harness.F.run --source "ground truth" --spatial-code-format explicit --scene SCENE
-python -m harness.F.launch --source perceived --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32
-python -m harness.F.sweep --sources all --spatial-code-formats all --depths metric --trackings tracking --input-selections uniform --frames 32
+python -m harness.F.run --source perceived --depth metric --tracking tracking --input-selection uniform --frames 32 --scene SCENE
+python -m harness.F.launch --source perceived --depth metric --tracking tracking --input-selection uniform --frames 32
+python -m harness.F.sweep --sources perceived --depths metric --trackings tracking --input-selections uniform --frames 32
+python -m harness.F.run --source perceived --depth metric --tracking tracking --video --scene SCENE
+python -m harness.F.launch --source perceived --depth metric --tracking tracking --video
+python -m harness.F.sweep --sources perceived --depths metric --trackings tracking --video
 ```
 
 Files:
@@ -346,39 +290,6 @@ Files:
 - `harness/F/launch.py`: thin launch entry point to `run.main`
 - `harness/F/sweep.py`: grid over symbolic source/config axes
 
-### Calibration
-
-```bash
-python -m calibration.run --models qwen3.5-4b,qwen3.5-2b --budgets 256,512,1024 --questions 1,2,3 --spatial-code-format explicit --depth metric --tracking tracking --input-selection uniform --frames 32
-python -m calibration.report --results-dir /root/results/calibration --tolerance 0.01 --max-forced-rate 0.05
-```
-
-Files:
-
-- `calibration/__init__.py`: calibration constants and result root
-- `calibration/run.py`: runs selected question ids across model/budget grid
-- `calibration/report.py`: summarizes calibration grid and recommends budget
-
-### Corruption Experiments
-
-```bash
-python -m corruption.sample --output /workspace/analysis/corruption_sample.json --seed 0
-python -m corruption.run --arm solver --transform position-jitter --magnitude 0.25 --spatial-code-format explicit --scenes scene1,scene2
-python -m corruption.run --arm vlm --model qwen3.5-4b --transform dimension-noise --magnitude 0.25 --spatial-code-format explicit --scenes scene1,scene2
-python -m corruption.launch --arm both --models qwen3.5-4b,internvl3.5-4b --transforms position-jitter,dimension-noise --magnitudes 0.1,0.25 --sample /workspace/analysis/corruption_sample.json
-python -m corruption.empirical --output /workspace/analysis/sam3_da3_residuals.json
-```
-
-Files:
-
-- `corruption/__init__.py`: corruption constants, transform names, result root
-- `corruption/transforms.py`: compact-code transformations
-- `corruption/chimera.py`: chimera and single-object intervention utilities
-- `corruption/empirical.py`: measured SAM3/DA3 residual distribution
-- `corruption/sample.py`: deterministic question sample builder
-- `corruption/run.py`: one corruption condition through solver and/or VLM
-- `corruption/launch.py`: grid launcher over transforms/magnitudes/models
-
 ### Analysis And Reports
 
 Current analysis is centered on report generation through `analysis.letters_reports` and
@@ -399,7 +310,6 @@ python -m analysis.letters_reports \
 python -m analysis.A_reports --results-dir /root/results/A --output-dir /workspace/reports
 python -m analysis.B_reports --results-dir /root/results/B --output-dir /workspace/reports
 python -m analysis.C_reports --results-dir /root/results/C --output-dir /workspace/reports
-python -m analysis.D_reports --results-dir /root/results/D --output-dir /workspace/reports
 python -m analysis.F_reports --results-dir /root/results/F --output-dir /workspace/reports
 ```
 
@@ -409,27 +319,20 @@ Files:
 - `analysis/A_reports.py`: wrapper for A report
 - `analysis/B_reports.py`: wrapper for B report
 - `analysis/C_reports.py`: wrapper for C report
-- `analysis/D_reports.py`: wrapper for D report
 - `analysis/F_reports.py`: wrapper for F report
 
 ## Result Layouts
 
 | Producer | Default output layout |
 |---|---|
-| Inference SAM3 | `/root/data/caches/sam3/<tracking>/<input>/<frames>/<scene>.pt` |
-| Inference DA3 | `/root/data/caches/depth-anything-3/<depth>/<input>/<frames>/<scene>.pkl` |
-| Encoder combined cache | `/root/data/caches/sam3+depth-anything-3/<depth>/<tracking>/<input>/<frames>/<scene>.pkl.gz` |
-| Perceived spatial code | `$VSI_CODES/sam3+depth-anything-3/<depth>/<tracking>/<input>/<frames>/<format>/<scene>.json` |
-| Overlay spatial code | `$VSI_CODES/overlay/sam3+depth-anything-3/<depth>/<tracking>/<input>/<frames>/explicit/<scene>.json` |
-| Ground-truth spatial code | `$VSI_CODES/ground truth/<format>/<scene>.json` |
-| A results | `/root/results/A/<model>/<protocol>/<selection>/<frames>/<scene>/<question_id>.json` |
-| B results | `/root/results/B/<model>/<protocol>/<format>/<depth>/<tracking>/<input>/<frames>/<scene>/<question_id>.json` |
-| C results | `/root/results/C[/overlay]/<model>/<protocol>/<format>/<depth>/<tracking>/<input>/<frames>/<scene>/<question_id>.json` |
-| D results | `/root/results/D/<model>/<input_mode>/<protocol>/<format>/.../<scene>/<question_id>.json` |
-| E results | `/root/results/E/<model>/<protocol>/<scene>/<question_id>.json` |
-| F results | `/root/results/F/<source>/.../<scene>/<question_id>.json` |
-| Corruption results | `/root/results/corruption/<arm>/<transform>/<magnitude>/<model-or-symbolic>/<scene>/<question_id>.json` |
-| Calibration results | `/root/results/calibration/...` |
+| Inference SAM3 | `/root/data/caches/sam3/<tracking>/frames/<input>/<frames>/<scene>.pt` |
+| Inference DA3 | `/root/data/caches/depth-anything-3/<depth>/frames/<input>/<frames>/<scene>.pkl` |
+| Encoder combined cache | `/root/data/caches/sam3+depth-anything-3/<tracking>/frames/<input>/<frames>/<scene>.pkl.gz` or `/root/data/caches/sam3+depth-anything-3/<tracking>/video/<scene>.pkl.gz` |
+| Perceived spatial code | `$VSI_CODES/sam3+depth-anything-3/<tracking>/frames/<input>/<frames>/explicit/<scene>.json` or `$VSI_CODES/sam3+depth-anything-3/<tracking>/video/explicit/<scene>.json` |
+| A results | `/root/results/A/<model>/{<selection>/<frames>|video}/<scene>/<question_id>.json` |
+| B results | `/root/results/B/<model>/explicit/<depth>/<tracking>/{<input>/<frames>|video}/<scene>/<question_id>.json` |
+| C results | `/root/results/C/<model>/explicit/<depth>/<tracking>/{<input>/<frames>|video}/<scene>/<question_id>.json` |
+| F results | `/root/results/F/perceived/<depth>/<tracking>/{<input>/<frames>|video}/explicit/<scene>/<question_id>.json` |
 | Reports | `/workspace/reports/*_report.json` |
 
 ## Full Source File Index
@@ -453,30 +356,8 @@ Files:
 | `analysis/A_reports.py` | CLI wrapper for A report |
 | `analysis/B_reports.py` | CLI wrapper for B report |
 | `analysis/C_reports.py` | CLI wrapper for C report |
-| `analysis/D_reports.py` | CLI wrapper for D report |
 | `analysis/F_reports.py` | CLI wrapper for F report |
 | `analysis/letters_reports.py` | Shared analysis/report implementation |
-| `analysis/letters_reports.py.orig` | Previous copy of the report module retained as a local reference artifact |
-
-### `calibration/`
-
-| File | Purpose |
-|---|---|
-| `calibration/__init__.py` | Calibration constants |
-| `calibration/run.py` | Run reasoning-budget pilots |
-| `calibration/report.py` | Summarize pilots and choose budget |
-
-### `corruption/`
-
-| File | Purpose |
-|---|---|
-| `corruption/__init__.py` | Corruption constants and transform vocabulary |
-| `corruption/transforms.py` | Compact-code perturbation functions |
-| `corruption/chimera.py` | Chimera/single-object interventions |
-| `corruption/empirical.py` | Measured residual distribution and empirical noise |
-| `corruption/sample.py` | Deterministic question sampler |
-| `corruption/run.py` | One corruption condition |
-| `corruption/launch.py` | Corruption grid launcher |
 
 ### `encoder/`
 
@@ -486,10 +367,9 @@ Files:
 | `encoder/config.py` | Paths and validated dimensions |
 | `encoder/adapters.py` | Raw model-cache adapters |
 | `encoder/geometric.py` | Spatial-code geometry math |
-| `encoder/render.py` | Writes compact/explicit JSON |
+| `encoder/render.py` | Writes explicit JSON |
 | `encoder/run.py` | One-scene perceived-code build |
 | `encoder/launch.py` | Batch perceived-code build |
-| `encoder/ground_truth.py` | Ground-truth-code build |
 
 ### `harness/`
 
@@ -511,23 +391,9 @@ Files:
 | `harness/B/sweep.py` | B sweep launcher |
 | `harness/C/__init__.py` | C constants |
 | `harness/C/prompts.py` | Frames+code prompts |
-| `harness/C/overlay.py` | Overlay frame/code artifacts |
-| `harness/C/overlay_launch.py` | Overlay pre-generation |
 | `harness/C/run.py` | C one-scene runner |
 | `harness/C/launch.py` | C batch launcher |
 | `harness/C/sweep.py` | C sweep launcher |
-| `harness/D/__init__.py` | D constants |
-| `harness/D/spatial_codes.py` | Ground-truth-code loader |
-| `harness/D/prompts.py` | Ground-truth-code prompts |
-| `harness/D/run.py` | D one-scene runner |
-| `harness/D/launch.py` | D batch launcher |
-| `harness/D/sweep.py` | D sweep launcher |
-| `harness/D/symbolic_eval.py` | Symbolic solver on ground-truth codes |
-| `harness/E/__init__.py` | E constants |
-| `harness/E/prompts.py` | Question-only prompts |
-| `harness/E/run.py` | E one-scene runner |
-| `harness/E/launch.py` | E batch launcher |
-| `harness/E/sweep.py` | E sweep launcher |
 | `harness/F/__init__.py` | F constants |
 | `harness/F/run.py` | Symbolic solver as harness |
 | `harness/F/launch.py` | Thin F launch entry point |
@@ -590,34 +456,10 @@ The tests mirror source folders. They are written to run without real data, resu
 - `tests/test_C/conftest.py`
 - `tests/test_C/test_C.py`
 - `tests/test_C/test_launch.py`
-- `tests/test_C/test_overlay.py`
-- `tests/test_C/test_overlay_launch.py`
 - `tests/test_C/test_prompts.py`
 - `tests/test_C/test_run.py`
 - `tests/test_C/test_sweep.py`
 
-### `tests/test_D/`
-
-- `tests/test_D/__init__.py`
-- `tests/test_D/conftest.py`
-- `tests/test_D/test_D.py`
-- `tests/test_D/test_launch.py`
-- `tests/test_D/test_prompts.py`
-- `tests/test_D/test_run.py`
-- `tests/test_D/test_spatial_codes.py`
-- `tests/test_D/test_sweep.py`
-- `tests/test_D/test_symbolic_eval.py`
-
-### `tests/test_E/`
-
-- `tests/test_E/__init__.py`
-- `tests/test_E/conftest.py`
-- `tests/test_E/test_E.py`
-- `tests/test_E/test_launch.py`
-- `tests/test_E/test_prompts.py`
-- `tests/test_E/test_run.py`
-- `tests/test_E/test_sweep.py`
-
 ### `tests/test_F/`
 
 - `tests/test_F/__init__.py`
@@ -634,7 +476,6 @@ The tests mirror source folders. They are written to run without real data, resu
 - `tests/test_analysis/test_A_reports.py`
 - `tests/test_analysis/test_B_reports.py`
 - `tests/test_analysis/test_C_reports.py`
-- `tests/test_analysis/test_D_reports.py`
 - `tests/test_analysis/test_F_reports.py`
 - `tests/test_analysis/test_analysis.py`
 - `tests/test_analysis/test_letters_reports.py`
@@ -643,26 +484,6 @@ The tests mirror source folders. They are written to run without real data, resu
 
 - `tests/test_backup.py`
 
-### `tests/test_calibration/`
-
-- `tests/test_calibration/__init__.py`
-- `tests/test_calibration/conftest.py`
-- `tests/test_calibration/test_calibration.py`
-- `tests/test_calibration/test_report.py`
-- `tests/test_calibration/test_run.py`
-
-### `tests/test_corruption/`
-
-- `tests/test_corruption/__init__.py`
-- `tests/test_corruption/conftest.py`
-- `tests/test_corruption/test_chimera.py`
-- `tests/test_corruption/test_corruption.py`
-- `tests/test_corruption/test_empirical.py`
-- `tests/test_corruption/test_launch.py`
-- `tests/test_corruption/test_run.py`
-- `tests/test_corruption/test_sample.py`
-- `tests/test_corruption/test_transforms.py`
-
 ### `tests/test_encoder/`
 
 - `tests/test_encoder/conftest.py`
@@ -718,8 +539,6 @@ When adding a Python source file, add the corresponding test in the mirrored tes
 folder. For example:
 
 ```text
-/workspace/harness/C/overlay.py
-/workspace/tests/test_C/test_overlay.py
 ```
 
 Every `tests/test_<folder>/` directory should contain both `conftest.py` and

backup.py

@@ -23,8 +23,6 @@ TARGETS = {
         "setup.sh",
         "backup.py",
         "analysis",
-        "calibration",
-        "corruption",
         "encoder",
         "harness",
         "inference",
@@ -38,11 +36,7 @@ TARGETS = {
     "A": ["results/A"],
     "B": ["results/B"],
     "C": ["results/C"],
-    "D": ["results/D"],
-    "E": ["results/E"],
     "F": ["results/F"],
-    "calibration": ["results/calibration"],
-    "corruption": ["results/corruption"],
 }
 
 

encoder/config.py

@@ -29,7 +29,9 @@ TRACKING_MODES = ("tracking", "no tracking")
 SPATIAL_CODE_FORMATS = ("compact", "explicit")
 
 
-def _validate_dimensions(depth, input_selection, tracking, frame_count, allow_video=False):
+def _validate_dimensions(
+    depth, input_selection, tracking, frame_count, allow_video=False
+):
     """Validate the dimensions shared by cache and spatial-code paths."""
     if depth not in DEPTH_VARIANTS:
         raise ValueError(f"unknown depth variant {depth!r}; expected {DEPTH_VARIANTS}")
@@ -42,15 +44,21 @@ def _validate_dimensions(depth, input_selection, tracking, frame_count, allow_vi
         raise ValueError(
             f"unknown tracking mode {tracking!r}; expected {TRACKING_MODES}"
         )
-    if frame_count < 1:
+    if input_selection != VIDEO_INPUT_SELECTION and (
+        frame_count is None or frame_count < 1
+    ):
         raise ValueError("frame count must be positive")
 
 
-def _combined_directory(root, depth, input_selection, tracking, frame_count, allow_video=False):
-    """Build the shared depth/tracking/input/frame-count hierarchy."""
+def _combined_directory(
+    root, depth, input_selection, tracking, frame_count, allow_video=False
+):
+    """Build the combined hierarchy: tracking/{frames/<selection>|video}."""
     _validate_dimensions(depth, input_selection, tracking, frame_count, allow_video)
-    directory = Path(root) / depth / tracking / input_selection
-    return directory if input_selection == VIDEO_INPUT_SELECTION else directory / str(frame_count)
+    directory = Path(root) / tracking
+    if input_selection == VIDEO_INPUT_SELECTION:
+        return directory / VIDEO_CACHE_SEGMENT
+    return directory / "frames" / input_selection / str(frame_count)
 
 
 def video_path(scene, dataset=None):
@@ -79,7 +87,12 @@ def video_path(scene, dataset=None):
 def cache_file(scene, depth, input_selection, tracking, frame_count=FRAMES_PER_VIDEO):
     """Return one combined geometry cache isolated by every input dimension."""
     directory = _combined_directory(
-        CACHE_ROOT / MODEL, depth, input_selection, tracking, frame_count, allow_video=True
+        CACHE_ROOT / MODEL,
+        depth,
+        input_selection,
+        tracking,
+        frame_count,
+        allow_video=True,
     )
     return str(directory / f"{scene}.pkl.gz")
 
@@ -87,31 +100,31 @@ def cache_file(scene, depth, input_selection, tracking, frame_count=FRAMES_PER_V
 def da3_cache_file(scene, depth, input_selection, frame_count=FRAMES_PER_VIDEO):
     """Return one native DA3 cache path; tracking never applies to DA3."""
     _validate_dimensions(depth, input_selection, "tracking", frame_count)
-    directory = CACHE_ROOT / "depth-anything-3" / depth / input_selection
+    directory = CACHE_ROOT / "depth-anything-3" / depth / "frames" / input_selection
     return str(directory / str(frame_count) / f"{scene}.pkl")
 
 
 def sam3_cache_file(scene, input_selection, tracking, frame_count=FRAMES_PER_VIDEO):
     """Return one native SAM3 cache path for a specific set of input dimensions."""
     _validate_dimensions("relative", input_selection, tracking, frame_count)
-    directory = CACHE_ROOT / "sam3" / tracking / input_selection
+    directory = CACHE_ROOT / "sam3" / tracking / "frames" / input_selection
     return str(directory / str(frame_count) / f"{scene}.pt")
 
 
 def video_da3_cache_file(scene, depth="metric", input_selection=VIDEO_FRAME_SELECTION):
     """Return the DA3 full-video cache for the encoder's video layout."""
     _validate_dimensions(depth, input_selection, "tracking", FRAMES_PER_VIDEO)
-    directory = CACHE_ROOT / "depth-anything-3" / depth / input_selection
-    return str(directory / VIDEO_CACHE_SEGMENT / f"{scene}.pkl")
+    directory = CACHE_ROOT / "depth-anything-3" / depth / VIDEO_CACHE_SEGMENT
+    return str(directory / f"{scene}.npz")
 
 
 def video_sam3_cache_file(
-    scene, tracking="tracking", input_selection=VIDEO_FRAME_SELECTION
+    scene, tracking="no tracking", input_selection=VIDEO_FRAME_SELECTION
 ):
     """Return the SAM3 full-video cache for the encoder's video layout."""
     _validate_dimensions("relative", input_selection, tracking, FRAMES_PER_VIDEO)
-    directory = CACHE_ROOT / "sam3" / tracking / input_selection
-    return str(directory / VIDEO_CACHE_SEGMENT / f"{scene}.pt")
+    directory = CACHE_ROOT / "sam3" / tracking / VIDEO_CACHE_SEGMENT
+    return str(directory / f"{scene}.pkl.gz")
 
 
 def spatial_code_model_dir(
@@ -128,7 +141,12 @@ def spatial_code_model_dir(
             f"expected {SPATIAL_CODE_FORMATS}"
         )
     directory = _combined_directory(
-        CODES_ROOT / MODEL, depth, input_selection, tracking, frame_count, allow_video=True
+        CODES_ROOT / MODEL,
+        depth,
+        input_selection,
+        tracking,
+        frame_count,
+        allow_video=True,
     )
     return str(directory / spatial_code_format)
 
@@ -146,25 +164,3 @@ def spatial_code_path(
         depth, input_selection, tracking, frame_count, spatial_code_format
     )
     return str(Path(directory) / f"{scene}.json")
-
-
-def ground_truth_spatial_code_dir(spatial_code_format="explicit"):
-    """Return the ground-truth spatial-code directory for one format.
-
-    Ground truth has no depth/tracking/input-selection/frame-count axis -- it is built once
-    per scene directly from the dataset's own 3D annotations (encoder.ground_truth), not from
-    a perception pipeline run under any particular config -- so it lives in its own top-level
-    "ground truth" segment rather than under MODEL's depth/tracking/input/frames hierarchy.
-    """
-    if spatial_code_format not in SPATIAL_CODE_FORMATS:
-        raise ValueError(
-            f"unknown spatial-code format {spatial_code_format!r}; "
-            f"expected {SPATIAL_CODE_FORMATS}"
-        )
-    return str(CODES_ROOT / "ground truth" / spatial_code_format)
-
-
-def ground_truth_spatial_code_path(scene, spatial_code_format="explicit"):
-    """Return the ground-truth spatial-code JSON path for one scene."""
-    directory = ground_truth_spatial_code_dir(spatial_code_format)
-    return str(Path(directory) / f"{scene}.json")

tests/test_F/test_F.py

@@ -7,7 +7,7 @@ from harness import F
 
 
 def test_sources_and_default_are_declared():
-    assert F.SOURCES == ("perceived", "ground truth")
+    assert F.SOURCES == ("perceived",)
     assert F.DEFAULT_SOURCE == "perceived"
     assert F.RESULTS_DIR == Path("/root/results/F")
 

tests/test_F/test_run.py

@@ -13,27 +13,7 @@ def test_perceived_layout_contains_every_input_axis():
     ) == Path("/root/results/F/perceived/metric/tracking/uniform/32/explicit")
 
 
-def test_ground_truth_layout_has_no_perception_axes():
-    assert run.results_dir_for("ground truth", "compact") == Path(
-        "/root/results/F/ground truth/compact"
-    )
-
-
-def test_select_source_routes_to_existing_symbolic_selectors(monkeypatch):
-    seen = []
-    monkeypatch.setattr(
-        run.symbolic_run, "select_spatial_codes", lambda *args: seen.append(("p", args))
-    )
-    monkeypatch.setattr(
-        run.symbolic_run,
-        "select_ground_truth_spatial_codes",
-        lambda *args: seen.append(("g", args)),
-    )
-    run.select_source(
-        "perceived", "compact", "relative", "no tracking", "selective", 64
-    )
-    run.select_source("ground truth", "explicit")
-    assert seen == [
-        ("p", ("relative", "selective", "no tracking", 64, "compact")),
-        ("g", ("explicit",)),
-    ]
+def test_video_results_use_video_branch():
+    assert run.results_dir_for(
+        "perceived", "explicit", "metric", "tracking", "video", None
+    ) == Path("/root/results/F/perceived/metric/tracking/video/explicit")

tests/test_analysis/test_analysis.py

@@ -10,7 +10,3 @@ class TestAnalysisDirectory(ReportTestCase):
         self.assertEqual(names, {"A_report.json", "C_report.json", "AC_report.json"})
         for name in names:
             json.loads((self.root / "reports" / name).read_text())
-
-    def test_E_is_excluded(self):
-        with self.assertRaises(ValueError):
-            load("E")

tests/test_encoder/test_config.py

@@ -9,26 +9,32 @@ def test_encoder_paths_mirror_all_dimensions(tmp_path, monkeypatch):
     monkeypatch.setattr(config, "CACHE_ROOT", tmp_path / "caches")
     monkeypatch.setattr(config, "CODES_ROOT", tmp_path / "codes")
     assert config.cache_file("scene", "metric", "uniform", "no tracking", 64).endswith(
-        "metric/no tracking/uniform/64/scene.pkl.gz"
+        "no tracking/frames/uniform/64/scene.pkl.gz"
     )
     assert config.da3_cache_file("scene", "relative", "uniform", 32).endswith(
-        "depth-anything-3/relative/uniform/32/scene.pkl"
+        "depth-anything-3/relative/frames/uniform/32/scene.pkl"
     )
     assert config.video_da3_cache_file("scene").endswith(
-        "depth-anything-3/metric/uniform/video/scene.pkl"
+        "depth-anything-3/metric/video/scene.npz"
+    )
+    assert config.video_da3_cache_file("scene", "relative").endswith(
+        "depth-anything-3/relative/video/scene.npz"
     )
     assert config.video_sam3_cache_file("scene").endswith(
-        "sam3/tracking/uniform/video/scene.pt"
+        "sam3/no tracking/video/scene.pkl.gz"
+    )
+    assert config.video_sam3_cache_file("scene", "tracking").endswith(
+        "sam3/tracking/video/scene.pkl.gz"
     )
     assert config.spatial_code_path(
         "scene", "metric", "selective", "tracking", 64
-    ).endswith("metric/tracking/selective/64/explicit/scene.json")
+    ).endswith("tracking/frames/selective/64/explicit/scene.json")
     assert config.spatial_code_path(
         "scene", "relative", "uniform", "no tracking", 96, "compact"
-    ).endswith("relative/no tracking/uniform/96/compact/scene.json")
+    ).endswith("no tracking/frames/uniform/96/compact/scene.json")
     assert config.spatial_code_path(
-        "scene", "metric", config.VIDEO_INPUT_SELECTION, "no tracking", 32, "explicit"
-    ).endswith("metric/no tracking/video/explicit/scene.json")
+        "scene", "metric", config.VIDEO_INPUT_SELECTION, "no tracking", None, "explicit"
+    ).endswith("no tracking/video/explicit/scene.json")
 
 
 def test_encoder_paths_reject_unknown_spatial_code_format():
@@ -36,20 +42,3 @@ def test_encoder_paths_reject_unknown_spatial_code_format():
         config.spatial_code_path(
             "scene", "metric", "uniform", "tracking", 32, "unknown"
         )
-
-
-def test_ground_truth_paths_have_no_depth_tracking_input_frames_axis(
-    tmp_path, monkeypatch
-):
-    monkeypatch.setattr(config, "CODES_ROOT", tmp_path / "codes")
-    assert config.ground_truth_spatial_code_path("scene", "explicit").endswith(
-        "codes/ground truth/explicit/scene.json"
-    )
-    assert config.ground_truth_spatial_code_path("scene", "compact").endswith(
-        "codes/ground truth/compact/scene.json"
-    )
-
-
-def test_ground_truth_paths_reject_unknown_spatial_code_format():
-    with pytest.raises(ValueError, match="unknown spatial-code format"):
-        config.ground_truth_spatial_code_path("scene", "unknown")

tests/test_encoder/test_render.py

@@ -6,4 +6,4 @@ from encoder import render
 def test_render_exposes_tracking_and_format_dimensions():
     variables = render.write_spatial_code_for.__code__.co_varnames
     assert "tracking" in variables
-    assert "spatial_code_format" in variables
+    assert "spatial_code_format" not in variables

tests/test_inference/test_inference.py

@@ -16,7 +16,7 @@ def test_output_paths_include_frame_hierarchy(tmp_path, monkeypatch):
     monkeypatch.setattr(run.inference_config, "CACHE_ROOT", tmp_path)
     assert run.output_path(
         "scene1", "SAM3", "uniform", "tracking", frame_count=64
-    ).endswith("sam3/tracking/uniform/64/scene1.pt")
+    ).endswith("sam3/tracking/frames/uniform/64/scene1.pt")
     assert run.output_path(
         "scene1", "DA3NESTED-GIANT-LARGE-1.1", "uniform", frame_count=64
-    ).endswith("depth-anything-3/metric/uniform/64/scene1.pkl")
+    ).endswith("depth-anything-3/metric/frames/uniform/64/scene1.pkl")

tests/test_inference/test_init.py

@@ -41,9 +41,9 @@ def test_cache_dir_helpers_validate_axes_and_include_dimensions(tmp_path, monkey
 
     assert inference.model_cache_dir(
         "depth-anything-3", "uniform", 16, "metric"
-    ) == str(tmp_path / "depth-anything-3" / "metric" / "uniform" / "16")
+    ) == str(tmp_path / "depth-anything-3" / "metric" / "frames" / "uniform" / "16")
     assert inference.sam3_cache_dir("no tracking", "selective", 8) == str(
-        tmp_path / "sam3" / "no tracking" / "selective" / "8"
+        tmp_path / "sam3" / "no tracking" / "frames" / "selective" / "8"
     )
     assert inference.parse_sam3_frame_mode("uniform-tracking") == (
         "uniform",

tests/test_symbolic/test_run.py

@@ -10,16 +10,25 @@ from symbolic import run
 def test_symbolic_selection_uses_new_hierarchy(tmp_path, monkeypatch):
     monkeypatch.setattr(run, "SPATIAL_CODES_ROOT", str(tmp_path))
     run.select_spatial_codes("metric", "uniform", "no tracking", 64, "explicit")
-    assert run.SPATIAL_CODES_DIR.endswith("metric/no tracking/uniform/64/explicit")
+    assert run.SPATIAL_CODES_DIR.endswith("no tracking/frames/uniform/64/explicit")
     assert run.results_dir_for_selection().endswith(
-        "metric/no tracking/uniform/64/explicit"
+        "no tracking/frames/uniform/64/explicit"
     )
 
 
+def test_symbolic_video_selection_uses_encoder_hierarchy(tmp_path, monkeypatch):
+    monkeypatch.setattr(run, "SPATIAL_CODES_ROOT", str(tmp_path))
+    run.select_spatial_codes("metric", "video", "no tracking", None, "explicit")
+    assert run.SPATIAL_CODES_DIR.endswith("no tracking/video/explicit")
+    assert run.results_dir_for_selection().endswith("no tracking/video/explicit")
+    assert run.SPATIAL_CODES_INPUT == "video"
+    assert run.SPATIAL_CODES_FRAMES is None
+
+
 def test_symbolic_selection_isolates_compact_codes(tmp_path, monkeypatch):
     monkeypatch.setattr(run, "SPATIAL_CODES_ROOT", str(tmp_path))
     run.select_spatial_codes("relative", "selective", "tracking", 96, "compact")
-    assert run.SPATIAL_CODES_DIR.endswith("relative/tracking/selective/96/compact")
+    assert run.SPATIAL_CODES_DIR.endswith("tracking/frames/selective/96/compact")
     assert run.SPATIAL_CODES_FORMAT == "compact"