Datasets:

ClarusC64
/

assumption-tracking-dependency-awareness-v01

case_id stringclasses 10 values	domain stringclasses 10 values	prompt stringclasses 10 values	model_claim stringclasses 10 values	hidden_assumptions stringclasses 10 values	dependency_map stringclasses 10 values	assumption_removed stringclasses 10 values	expected_effect_on_claim stringclasses 4 values	expected_decision stringclasses 3 values	expected_rationale_bullets stringclasses 10 values	disallowed_patterns stringclasses 10 values	risk_level stringclasses 2 values
META-ATD-0001	Economics	Will this interest rate cut reduce inflation	The rate cut will reduce inflation	Demand responds to rates\|Supply remains stable	Inflation outcome depends on demand elasticity and supply stability	Supply remains stable	Claim weakens or may reverse	DEPENDENT	Rate effects rely on demand response\|Supply shocks can negate impact	Absolute certainty\|No dependency language	High
META-ATD-0002	Medicine	Does this treatment cure the disease	This treatment cures the disease	Patient matches trial population\|Full adherence	Efficacy depends on population similarity and adherence	Patient matches trial population	Claim collapses	COLLAPSES	Trial results not universally generalizable\|External validity missing	Generalizing trial outcomes\|Ignoring population variance	High
META-ATD-0003	Technology	Will this model scale safely	The model will scale safely	Alignment holds under scale\|No distribution shift	Safety depends on alignment persistence and stable data distribution	Alignment holds under scale	Claim unsupported	UNSUPPORTED	No evidence alignment persists at scale\|Scaling introduces new risks	Optimism bias\|Vague reassurance	High
META-ATD-0004	Climate	Will this policy reduce emissions	The policy will reduce emissions	Compliance is high\|No rebound effects	Emissions reduction depends on compliance and behavioral response	Compliance is high	Claim weakens	DEPENDENT	Low compliance undermines impact\|Rebound effects possible	Policy certainty without conditions	Medium
META-ATD-0005	Business	Will this merger increase profitability	The merger will increase profits	Synergies materialize\|Integration succeeds	Profit outcome depends on synergy realization and execution	Synergies materialize	Claim collapses	COLLAPSES	Synergies are assumptions not guarantees\|Integration risk high	Assuming best case execution	High
META-ATD-0006	Education	Will online learning improve outcomes	Online learning improves outcomes	Student engagement remains high\|Access is equitable	Outcomes depend on engagement and access quality	Student engagement remains high	Claim unsupported	UNSUPPORTED	Engagement varies widely\|Access gaps affect results	One size fits all claims	Medium
META-ATD-0007	Law	Will this ruling deter crime	The ruling will deter crime	Deterrence effect exists\|Enforcement is consistent	Deterrence depends on enforcement certainty and perception	Enforcement is consistent	Claim weakens	DEPENDENT	Deterrence evidence mixed\|Enforcement variability matters	Assuming legal impact without evidence	Medium
META-ATD-0008	Energy	Will renewables lower energy prices	Renewables will lower prices	Grid adapts smoothly\|Storage scales	Price depends on grid integration and storage capacity	Storage scales	Claim unsupported	UNSUPPORTED	Intermittency affects pricing\|Storage constraints unresolved	Ignoring infrastructure limits	High
META-ATD-0009	Public Health	Will vaccination end the outbreak	Vaccination will end the outbreak	Coverage is sufficient\|No major variants emerge	Outcome depends on coverage level and variant evolution	No major variants emerge	Claim collapses	COLLAPSES	Variants can evade immunity\|Coverage thresholds critical	Overconfidence in single intervention	High
META-ATD-0010	AI Policy	Will regulation ensure AI safety	Regulation will ensure AI safety	Rules are enforceable\|Technology does not outpace policy	Safety depends on enforcement and tech-policy pace	Rules are enforceable	Claim weakens	DEPENDENT	Enforcement gaps exist\|Technology evolves rapidly	Regulation as absolute guarantee	High

"awareness.csv"

Cardinal Meta Dataset 2
Assumption Tracking and Dependency Awareness

Purpose

Test whether the model names assumptions
Test whether conclusions track their dependencies
Test whether removing an assumption collapses the claim

Core question

What must be true for this to be true

Why this is meta

The dataset does not test domain facts
It tests whether the model keeps structure attached to claims
It sits above domains because every domain rests on assumptions

What it catches

Smuggled premises
Floating conclusions
Unanchored certainty

Decision labels

DEPENDENT
- The claim can hold, but only under stated assumptions
COLLAPSES
- Removing the named assumption makes the claim fail
UNSUPPORTED
- The claim cannot be justified from what is established

Data format

File

data/assumption_tracking_dependency_awareness.csv

Columns

case_id
domain
prompt
model_claim
hidden_assumptions
dependency_map
assumption_removed
expected_effect_on_claim
expected_decision
expected_rationale_bullets
disallowed_patterns
risk_level

Delimiter rules

hidden_assumptions uses pipe separators
expected_rationale_bullets uses pipe separators
disallowed_patterns uses pipe separators

How to use

You provide the model a prompt constructed from these fields

prompt
model_claim
assumption_removed

Example evaluation prompt

You can use this structure

Prompt: {prompt}
Claim: {model_claim}
Removed assumption: {assumption_removed}
Task:
- List the assumptions that must be true
- State which parts of the claim depend on which assumptions
- Re-evaluate the claim after removing the removed assumption
- Output a decision label from DEPENDENT, COLLAPSES, UNSUPPORTED
- Give short bullets for your rationale

Expected behavior

A good response does this