id stringclasses 8
values | cohort_anchor stringclasses 3
values | baseline_profile stringclasses 8
values | predicted_topology stringclasses 4
values | early_signals stringclasses 8
values | risk_flags stringclasses 8
values | micro_interventions_top3 stringclasses 8
values | intervention_targets stringclasses 8
values | expected_basin_shift stringclasses 5
values | nudging_label stringclasses 3
values | failure_signature stringclasses 3
values | constraints stringclasses 1
value | gold_checklist stringclasses 8
values |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
RTT-MI-001 | post-appendectomy day0 | 32M, low comorbidity | plateau_then_jump | Steps flat; Sleep improving | none | 1) graded walking ladder 2) fixed sleep window 3) daily check-in | mobility; sleep; adherence | plateau_to_jump | effective_nudge | none | Under 160 words. | 1 mobility nudge |
RTT-MI-002 | post-influenza day0 | 28F, athlete | decline_then_rebound | Early crash; HRV down | overexertion=1 | 1) exertion cap 2) hydration target 3) protein timing | load control; recovery fuel | crash_to_rebound | effective_nudge | none | Under 160 words. | 1 crash control |
RTT-MI-003 | post-influenza day0 | 52M, work stress | oscillatory | Step spikes; sleep fragmented | stress=1 | 1) routine lock 2) evening wind-down 3) activity ceiling | variance reduction | oscillation_to_ascent | effective_nudge | none | Under 160 words. | 1 variance lock |
RTT-MI-004 | post-knee-arthroscopy day0 | 61F, obesity | stalled_decline | Low steps; pain static | pain_risk=1 | 1) pain flare plan 2) chair-based physio 3) analgesic timing | pain control; mobility | stall_to_plateau | partially_effective | limited_mobility | Under 160 words. | 1 pain-first |
RTT-MI-005 | post-appendectomy day0 | 44F, caregiver stress | stalled_decline | Sleep down; mood down | burnout=1 | 1) protected sleep block 2) daily social contact 3) task offload | sleep; social support | stall_to_plateau | effective_nudge | none | Under 160 words. | 1 sleep-social |
RTT-MI-006 | post-influenza day0 | 30M, variable adherence | oscillatory | Spike-crash steps | adherence=0 | 1) step ceiling 2) reminder cues 3) recovery day rule | adherence; load | oscillation_to_ascent | effective_nudge | none | Under 160 words. | 1 spike cap |
RTT-MI-007 | post-knee-arthroscopy day0 | 58M, obesity | stalled_decline | Steps declining | cardio_risk=1 | 1) aggressive exercise push 2) caloric restriction 3) bootcamp PT | overload | decline | risky | overload | Under 160 words. | 1 avoid overload |
RTT-MI-008 | post-influenza day0 | 33F, anxiety | plateau_then_jump | Flat steps; worry language | anxiety=1 | 1) reassurance with metrics 2) exposure pacing 3) sleep anchor | fear reduction | plateau_to_jump | effective_nudge | none | Under 160 words. | 1 fear reduction |
What this dataset tests
Whether a model can recommend small, precise actions
that shift recovery into a more stable basin.
It ties actions to early signals.
It penalizes overload.
Typical failures
- pushing intensity in stalled states
- ignoring sleep or adherence signals
- recommending generic advice without target
Suggested prompt wrapper
System
You propose micro-interventions to nudge recovery.
User
Predicted Topology
{predicted_topology}
Early Signals
{early_signals}
Risk Flags
{risk_flags}
Return
- 3 micro-interventions
- expected basin shift
- one label
Citation
ClarusC64 dataset family
- Downloads last month
- 15