Skip to content

task_charts__curve_panels__cross_panel_delta_extremum_label

Contract

  1. Domain: charts
  2. Scene id: curve_panels
  3. Source implementation domain/group: charts/curve_panels
  4. Query id: single
  5. Semantic query details are recorded in query_id and trace params.
  6. Controlled-unanswerable sampling may request a visible method label that is missing from one subplot.

Implementation

  1. Registered class: trace_tasks.tasks.charts.curve_panels.cross_panel_delta_extremum_label.ChartsScientificCrossPanelDeltaExtremumLabelTask
  2. Prompt lookup domain/group: charts/curve_panels
  3. Generation is deterministic from instance_seed, explicit params, prompt bundle, renderer config, and code versions.
  4. Answers and annotation are produced from the same metadata execution trace.

Annotation Contract

  1. Answer schema: string_label or the literal string unanswerable.
  2. Annotation schema: point_map.
  3. Annotation should mark the minimal visual witnesses required by the task, following the cross-domain annotation policy.
  4. Answerable instances map start_point and end_point to points on the selected method markers in the answer subplot.
  5. Unanswerable instances use an empty annotation object because the requested method is not plotted in every subplot.
  6. Renderer context such as legends, axes, decorative labels, titles, and distractor text is metadata unless the task explicitly asks for it as annotation.

Program Contract

Program: if plotted_in_every_panel(method): argmax(panel_label, value(method,end_x,panel_label)-value(method,start_x,panel_label)); else unanswerable; output=string_label|unanswerable; annotation=point_map(start_point,end_point)|empty_map; scene=curve_panels; scope=cross_panel_delta_extremum_label

Candidate set: the visible curve panels, curve traces, points, and panel labels inside the cross_panel_delta_extremum_label objective scope. Operands: prompt-bound labels, categories, series names, thresholds, intervals, references, and encoded chart values, plus the task's prompt-bound target operands when present. Operation: evaluate if plotted_in_every_panel over the candidate set using the filters, comparisons, aggregations, rankings, projections, or counterfactual edits named in the program expression; generation enforces a unique final answer. Output binding: answer is the string_label value bound by string_label|unanswerable. Annotation witnesses: point_map witnesses bound by point_map(start_point,end_point)|empty_map. Annotation should mark the minimal visual witnesses required by the task, following the cross-domain annotation policy. Answerable instances map start_point and end_point to points on the selected method markers in the answer subplot. Unanswerable instances use an empty annotation object because the requested method is not plotted in every subplot. Query ids: single.

Reasoning Operations

Families: ranking, formula_evaluation

Query Details

Query id Program signature Answer schema Annotation schema
single selection.cross_panel_delta_extremum_label string_label_or_unanswerable point_map