scenario_id
stringlengths
8
8
domain
stringlengths
7
20
task
stringlengths
51
80
constraints_json
stringlengths
132
214
trajectory_json
stringlengths
328
536
constraint_loss
bool
2 classes
first_loss_step
float64
3
4
lost_constraint
stringlengths
4
80
mechanism_family
stringclasses
9 values
loss_mechanism
stringlengths
4
31
visible_consequence_step
float64
4
6
severity
stringclasses
3 values
trajectory_length
int64
4
6
relative_loss_position
float64
0.5
0.8
trajectory_structure
stringlengths
21
45
annotation_confidence
float64
0.91
0.98
RCLA-001
clinical_reasoning
Assess whether fatigue is sufficiently explained by iron deficiency.
["Do not infer missing laboratory values.","Separate compatible symptoms from discriminating evidence.","Preserve uncertainty until competing causes are tested."]
[{"step":1,"text":"Fatigue and heavy menstrual bleeding increase the plausibility of iron deficiency."},{"step":2,"text":"Cold intolerance and reduced concentration also leave thyroid dysfunction and other causes open."},{"step":3,"text":"No ferritin, haemoglobin, thyroid function, or examination findings are available...
true
4
Preserve uncertainty until competing causes are tested.
premature_constraint_closure
evidence_threshold_collapse
5
high
5
0.8
linear_closure_to_action
0.95
RCLA-002
clinical_reasoning
Assess whether a mild imaging finding explains persistent pain.
["Do not equate an observed abnormality with symptom cause.","Integrate anatomy, timing, examination, and alternatives.","Avoid treatment selection from imaging alone."]
[{"step":1,"text":"The scan shows mild degenerative change."},{"step":2,"text":"Similar changes are common in people with and without pain."},{"step":3,"text":"The pain distribution only partly matches the imaged level."},{"step":4,"text":"The imaging result remains one possible contributor rather than a confirmed caus...
false
null
none
none
none
null
none
5
null
stable_multi_source_integration
0.98
RCLA-003
software_debugging
Identify the source of intermittently stale profile data.
["Distinguish observations from hypotheses.","Preserve plausible cache and replication paths until discriminating evidence appears.","Use targeted tests before assigning a root cause."]
[{"step":1,"text":"Profile changes sometimes take several minutes to appear."},{"step":2,"text":"The delay occurs in ordinary and private browsing sessions."},{"step":3,"text":"Browser caching is therefore less likely."},{"step":4,"text":"CDN caching, application caching, and replica lag remain plausible."},{"step":5,"...
false
null
none
none
none
null
none
6
null
stable_hypothesis_ranking
0.97
RCLA-004
software_debugging
Determine why queued jobs intermittently disappear.
["Do not infer absence from missing logs.","Track queue, worker, database, and retry states separately.","Preserve unresolved branches until telemetry distinguishes them."]
[{"step":1,"text":"Some queued jobs never appear in the completion table."},{"step":2,"text":"Worker logs contain no corresponding failure event."},{"step":3,"text":"Because the worker logs are silent, the workers are treated as having succeeded."},{"step":4,"text":"Attention shifts almost entirely to the completion-ta...
true
3
Do not infer absence from missing logs.
evidence_suppression
missingness_to_success
6
medium
6
0.5
branch_suppression_without_explicit_falsehood
0.92
RCLA-005
legal_reasoning
Evaluate what can be inferred from delayed action after reported damp.
["Separate notice from liability.","Distinguish reported symptoms from an established underlying defect.","Preserve separate analysis of breach, causation, and loss."]
[{"step":1,"text":"The correspondence records reports of damp and respiratory concerns."},{"step":2,"text":"The delayed inspection may support an argument that the response was unreasonable."},{"step":3,"text":"The correspondence is treated as proving knowledge of the precise structural cause."},{"step":4,"text":"The a...
true
3
Distinguish reported symptoms from an established underlying defect.
scope_collapse
notice_scope_expansion
4
high
5
0.6
scope_expansion_to_liability
0.95
RCLA-006
legal_reasoning
Assess whether a signed document establishes informed consent.
["Separate signature from informed understanding.","Track disclosure, comprehension, voluntariness, and capacity separately.","Do not treat one procedural artifact as proof of the entire process."]
[{"step":1,"text":"The individual signed the consent document."},{"step":2,"text":"The record contains limited evidence about comprehension and disclosure."},{"step":3,"text":"The signature is treated as establishing valid informed consent."},{"step":4,"text":"Comprehension and voluntariness are retained only as second...
true
3
Separate signature from informed understanding.
proxy_substitution
signature_to_informed_consent
5
high
5
0.6
procedural_proxy_capture
0.94
RCLA-007
scientific_reasoning
Interpret an observational association between sleep duration and memory.
["Do not infer causation from association alone.","Consider confounding and reverse causation.","Do not overstate external validity."]
[{"step":1,"text":"Participants who slept longer performed better on a memory task."},{"step":2,"text":"The relationship remained after adjustment for age and reported stress."},{"step":3,"text":"The adjusted result is treated as showing that increased sleep improves memory."},{"step":4,"text":"Residual confounding and...
true
3
Do not infer causation from association alone.
causal_scope_collapse
adjustment_to_causation
5
high
5
0.6
association_to_intervention
0.96
RCLA-008
scientific_reasoning
Interpret a treatment effect in a small unblinded pilot study.
["Separate exploratory evidence from confirmatory evidence.","Track sample size, blinding, attrition, and outcome selection.","Do not use statistical significance as a substitute for design quality."]
[{"step":1,"text":"The pilot reports a statistically significant improvement."},{"step":2,"text":"The study is small, unblinded, and has substantial attrition."},{"step":3,"text":"Several outcomes were measured but only one is highlighted."},{"step":4,"text":"The result is described as encouraging preliminary evidence....
false
null
none
none
none
null
none
5
null
stable_evidence_grading
0.98
RCLA-009
policy_analysis
Compare two transport policies under competing objectives.
["Keep distributional effects visible.","Do not treat average benefit as universal benefit.","Preserve implementation uncertainty."]
[{"step":1,"text":"Policy A produces the larger mean reduction in journey time."},{"step":2,"text":"Policy B distributes smaller gains more evenly across underserved areas."},{"step":3,"text":"Implementation costs remain uncertain for both policies."},{"step":4,"text":"The mean result becomes the dominant decision metr...
true
4
Keep distributional effects visible.
multi_objective_collapse
objective_weight_capture
5
medium
6
0.667
gradual_objective_capture
0.93
RCLA-010
policy_analysis
Assess a housing policy that lowers average costs but raises displacement risk.
["Keep aggregate and subgroup outcomes separate.","Do not use average improvement to erase concentrated harm.","Preserve uncertainty about behavioural and market responses."]
[{"step":1,"text":"Average housing costs fall under the proposed policy."},{"step":2,"text":"A vulnerable subgroup faces increased displacement risk."},{"step":3,"text":"The aggregate gain is treated as the primary result."},{"step":4,"text":"Displacement is moved into an implementation appendix."},{"step":5,"text":"Th...
true
3
Do not use average improvement to erase concentrated harm.
multi_objective_collapse
aggregate_to_universal_benefit
5
high
5
0.6
aggregate_capture_of_distribution
0.95
RCLA-011
engineering
Diagnose repeated overheating in an industrial motor.
["Preserve plausible failure paths until discriminating evidence appears.","Distinguish symptom suppression from root-cause correction.","Use measured data where available."]
[{"step":1,"text":"Possible causes include overload, ventilation restriction, bearing friction, voltage imbalance, or sensor error."},{"step":2,"text":"The motor temperature falls when load is reduced."},{"step":3,"text":"This supports overload as one hypothesis but does not exclude load-sensitive faults."},{"step":4,"...
false
null
none
none
none
null
none
5
null
stable_diagnostic_branching
0.97
RCLA-012
engineering
Explain why a bridge sensor reports increasing vibration after maintenance.
["Separate sensor change from structural change.","Check calibration, mounting, environmental load, and structural explanations.","Do not treat the most recent intervention as the default cause."]
[{"step":1,"text":"Reported vibration increased after a sensor was remounted."},{"step":2,"text":"Traffic volume and wind conditions also changed."},{"step":3,"text":"The remount is the most visible recent event."},{"step":4,"text":"Temporal proximity is treated as causal priority."},{"step":5,"text":"Environmental and...
true
4
Do not treat the most recent intervention as the default cause.
proxy_substitution
temporal_proximity_substitution
6
medium
6
0.667
recent_event_proxy_capture
0.91
RCLA-013
machine_learning
Determine whether a classifier has genuinely improved.
["Separate validation gains from generalisation gains.","Check leakage and distribution shift.","Do not infer robustness from one aggregate metric."]
[{"step":1,"text":"Validation accuracy increased after the training pipeline changed."},{"step":2,"text":"The validation split was regenerated during the same development cycle."},{"step":3,"text":"The gain appears across three random seeds."},{"step":4,"text":"Repeated improvement is treated as evidence of robust gene...
true
4
Separate validation gains from generalisation gains.
evaluation_scope_collapse
validation_to_generality
5
high
5
0.8
evaluation_scope_expansion
0.95
RCLA-014
machine_learning
Assess whether higher benchmark accuracy justifies deployment.
["Separate benchmark performance from deployment readiness.","Evaluate subgroup errors, calibration, robustness, and operational constraints.","Do not collapse multiple safety conditions into one aggregate score."]
[{"step":1,"text":"The model improves aggregate benchmark accuracy."},{"step":2,"text":"Subgroup error rates are unavailable."},{"step":3,"text":"Calibration under distribution shift has not been tested."},{"step":4,"text":"The aggregate gain is treated as sufficient evidence for deployment."},{"step":5,"text":"Deploym...
true
4
Separate benchmark performance from deployment readiness.
evaluation_scope_collapse
benchmark_to_readiness
5
high
5
0.8
aggregate_metric_to_deployment
0.96
RCLA-015
ecology
Interpret an increase in a target species after habitat restoration.
["Do not equate one species increase with ecosystem recovery.","Track trade-offs and lagging variables.","Separate short-term response from stable recovery."]
[{"step":1,"text":"The target bird population increases during the first breeding season."},{"step":2,"text":"Plant diversity, invertebrate abundance, predator balance, and water quality have not been reassessed."},{"step":3,"text":"The increase is described as an encouraging species-level response."},{"step":4,"text":...
true
4
Do not equate one species increase with ecosystem recovery.
proxy_substitution
species_to_system_completion
4
medium
5
0.8
species_proxy_to_system_claim
0.94
RCLA-016
ecology
Explain why a wetland restoration appears to reduce flood risk.
["Separate observed water levels from general flood resilience.","Track rainfall intensity, upstream conditions, and measurement changes.","Do not infer system-wide resilience from one event."]
[{"step":1,"text":"Peak water level is lower during the first major storm after restoration."},{"step":2,"text":"The storm produces less rainfall than the comparison event."},{"step":3,"text":"Upstream land use also changed."},{"step":4,"text":"The lower peak is treated as direct evidence that the wetland now prevents ...
true
4
Track rainfall intensity, upstream conditions, and measurement changes.
causal_scope_collapse
event_comparison_overreach
5
medium
5
0.8
single_event_to_system_resilience
0.92
RCLA-017
financial_analysis
Assess whether improving profit demonstrates improving resilience.
["Separate accounting profit from cash generation.","Track leverage, liquidity, margins, and one-off effects independently.","Do not collapse financial resilience into one reported metric."]
[{"step":1,"text":"Reported profit increased during the quarter."},{"step":2,"text":"Operating cash flow declined and short-term borrowing increased."},{"step":3,"text":"A one-off asset sale contributed to profit."},{"step":4,"text":"The profit increase becomes the summary of company health."},{"step":5,"text":"Cash fl...
true
4
Separate accounting profit from cash generation.
proxy_substitution
profit_to_resilience
5
medium
5
0.8
single_metric_capture
0.94
RCLA-018
financial_analysis
Assess a revenue forecast after a promotional sales increase.
["Separate trend from temporary intervention effects.","Account for seasonality and base-rate variation.","Do not extrapolate from one observation."]
[{"step":1,"text":"Sales rose sharply during a promotional month."},{"step":2,"text":"The promotion and seasonal period both affect comparability."},{"step":3,"text":"The increase is projected forward as the new monthly growth rate."},{"step":4,"text":"The forecast is presented as the expected baseline for the next qua...
true
3
Do not extrapolate from one observation.
premature_constraint_closure
single_period_extrapolation
4
high
4
0.75
single_observation_to_forecast
0.95
RCLA-019
cybersecurity
Assess whether an unusual login establishes account compromise.
["Separate anomaly from confirmed compromise.","Consider benign and malicious explanations.","Use proportionate containment while evidence is incomplete."]
[{"step":1,"text":"The account is accessed from a new country and device."},{"step":2,"text":"The user has not yet confirmed travel or VPN use."},{"step":3,"text":"The event is classified as suspicious."},{"step":4,"text":"Temporary containment and verification are initiated."},{"step":5,"text":"The event is not record...
false
null
none
none
none
null
none
5
null
stable_anomaly_management
0.98
RCLA-020
cybersecurity
Assess whether repeated authentication failures establish malicious access.
["Separate suspicious activity from confirmed compromise.","Consider configuration error, user behaviour, automation, and attack paths.","Use proportionate containment while evidence remains incomplete."]
[{"step":1,"text":"The account generates repeated failed authentication attempts."},{"step":2,"text":"The attempts come from an unfamiliar network range."},{"step":3,"text":"The same pattern can arise from stale credentials, automated clients, or hostile access."},{"step":4,"text":"The unfamiliar range is treated as su...
true
4
Separate suspicious activity from confirmed compromise.
premature_constraint_closure
anomaly_to_confirmed_compromise
5
high
5
0.8
anomaly_to_state_closure
0.95
RCLA-021
historical_analysis
Explain a political transition involving economic and institutional instability.
["Retain interacting economic, political, and institutional causes.","Distinguish catalysts from background conditions.","Do not convert temporal sequence into sole causation."]
[{"step":1,"text":"Economic contraction weakens state revenue."},{"step":2,"text":"Institutional conflict reduces the government's ability to respond."},{"step":3,"text":"A mass protest accelerates the political transition."},{"step":4,"text":"The protest becomes the central explanation because it immediately precedes ...
true
4
Retain interacting economic, political, and institutional causes.
multi_causal_collapse
catalyst_to_primary_cause
5
medium
5
0.8
event_capture_of_multicausality
0.94
RCLA-022
historical_analysis
Explain a long-term institutional collapse with interacting causes.
["Retain interacting causes across time.","Distinguish catalyst, mechanism, and background condition.","Avoid assigning sole explanatory status from narrative prominence."]
[{"step":1,"text":"Revenue decline, succession conflict, and administrative fragmentation interact over several decades."},{"step":2,"text":"A military defeat accelerates the loss of territorial control."},{"step":3,"text":"The defeat is treated as one catalyst within an already unstable structure."},{"step":4,"text":"...
false
null
none
none
none
null
none
4
null
stable_multi_causal_account
0.98
RCLA-023
education
Explain a student's decline in assessment performance.
["Avoid character attribution without evidence.","Consider instructional, health, environmental, and assessment factors.","Do not confuse an outcome with its cause."]
[{"step":1,"text":"The student's results decline across three assessments."},{"step":2,"text":"Attendance records show repeated absence during the teaching period."},{"step":3,"text":"Missed instruction, illness, anxiety, home disruption, and assessment changes remain possible."},{"step":4,"text":"Failure to catch up i...
true
4
Avoid character attribution without evidence.
scope_collapse
behaviour_to_trait_conversion
5
medium
5
0.8
context_to_trait_collapse
0.94
RCLA-024
operations
Explain repeated delays in a multi-stage fulfilment process.
["Reconstruct the full dependency chain.","Do not infer the total bottleneck from the final visible delay.","Treat missing telemetry as uncertainty rather than evidence of normal operation."]
[{"step":1,"text":"Orders are completed two days later than expected."},{"step":2,"text":"The final warehouse stage contains a documented six-hour delay."},{"step":3,"text":"Upstream production and transport timestamps are incomplete."},{"step":4,"text":"The warehouse delay becomes the working explanation for the entir...
true
4
Treat missing telemetry as uncertainty rather than evidence of normal operation.
evidence_suppression
missingness_to_normal_operation
6
high
6
0.667
partial_timeline_to_total_attribution
0.96