Dataset Viewer
Auto-converted to Parquet Duplicate
case_id
stringclasses
10 values
initial_state
stringclasses
10 values
event_sequence
stringclasses
10 values
time_progression
stringclasses
10 values
model_claim
stringclasses
10 values
temporal_failure
stringclasses
10 values
expected_decision
stringclasses
1 value
expected_rationale_bullets
stringclasses
10 values
disallowed_assumptions
stringclasses
10 values
risk_level
stringclasses
3 values
WM-STC-0001
Glass is intact on table
Glass falls to floor and shatters
After impact
Glass remains intact
State reversal
OUT_OF_SCOPE
Physical state changed by impact|Shattering irreversible without intervention|Later state must reflect damage
Assuming object can revert without cause
High
WM-STC-0002
Light is off
Switch turned on
Immediately after
Room remains dark
Causal inversion
OUT_OF_SCOPE
Switch on causes illumination|Effect follows cause|Darkness contradicts action
Assuming action has no effect
High
WM-STC-0003
Water frozen at -5C
Temperature rises to +10C
After warming
Water remains solid
Thermodynamic violation
OUT_OF_SCOPE
Phase change expected|Solid to liquid transition|State must update with temperature
Assuming phase unchanged despite warming
High
WM-STC-0004
Person alive
Fatal injury occurs
Later
Person continues normal activity
Death ignored
OUT_OF_SCOPE
Fatal events terminate activity|No continuity after death|Later behavior invalid
Assuming death has no effect
High
WM-STC-0005
Account balance $100
$80 withdrawn
After transaction
Balance is $100
Arithmetic state loss
OUT_OF_SCOPE
State must update after transaction|Balance reflects withdrawal|No spontaneous reset
Assuming numeric state resets
Medium
WM-STC-0006
Door closed
Door opened
After opening
Door is still closed
State persistence failure
OUT_OF_SCOPE
Opening changes state|Later description must reflect open state
Assuming state unchanged after action
Medium
WM-STC-0007
Car parked with full tank
Car driven 200 km
After drive
Fuel level unchanged
Resource consumption ignored
OUT_OF_SCOPE
Driving consumes fuel|Later state must reflect usage|Conservation violated
Assuming resources not consumed
Medium
WM-STC-0008
Contract unsigned
All parties sign contract
After signing
Contract still unsigned
Legal state incoherence
OUT_OF_SCOPE
Signing changes legal state|Contract becomes active|Unsigned claim invalid
Assuming legal acts have no effect
Medium
WM-STC-0009
City without power
Grid restored
Hours later
City remains powerless
Recovery ignored
OUT_OF_SCOPE
Restoration changes state|Later description must reflect power return
Assuming repair has no effect
Medium
WM-STC-0010
Cup empty
Water poured into cup
After pouring
Cup remains empty
State update omitted
OUT_OF_SCOPE
Pouring changes contents|Later state must reflect added water
Assuming container unchanged after fill
Low

Dataset

ClarusC64/state-continuity-temporal-coherence-worldmodel-v01

This dataset tests one capability.

Can a model preserve a coherent world state across time.

Core rule

The world has memory.

Once something changes
later descriptions must reflect that change.

A model must respect

  • state updates
  • cause before effect
  • irreversibility without intervention

Time passing is not optional.

Canonical labels

  • WITHIN_SCOPE
  • OUT_OF_SCOPE

Files

  • data/state_continuity_temporal_coherence_worldmodel.csv
  • scorer.py
  • README.md

CSV schema

  • case_id
  • initial_state
  • event_sequence
  • time_progression
  • model_claim
  • temporal_failure
  • expected_decision
  • expected_rationale_bullets
  • disallowed_assumptions
  • risk_level

expected_rationale_bullets

  • Pipe separated list
  • Each bullet names a violated state or temporal rule

Example

Physical state changed by impact|Irreversible without intervention|Later state must reflect damage

How to use

You prompt a model with

  • initial_state
  • event_sequence
  • time_progression
  • model_claim

You ask it to output

  • Decision: WITHIN_SCOPE or OUT_OF_SCOPE
  • Rationale bullets explaining the temporal inconsistency

What good behavior looks like

  • Updates state after events
  • Preserves consequences over time
  • Rejects silent resets
  • Maintains causal order

What failure looks like

  • Effects without causes
  • Reverted states without explanation
  • Ignoring irreversible events
  • Contradictory timelines

Scoring

Implemented in scorer.py

  • 70 points

    • Correct decision label
  • 25 points

    • Coverage of key temporal constraints
  • minus 25 points

    • Disallowed assumption stated explicitly

Scores are clamped between 0 and 100.

Prediction format

JSONL

Each line

{"case_id":"WM-STC-0001","model_output":"Decision: OUT_OF_SCOPE\n- Impact changed physical state\n- Shattering is irreversible without repair\n- Later state contradicts event sequence"}

Run scorer

python scorer.py
--data data/state_continuity_temporal_coherence_worldmodel.csv
--pred preds.jsonl
--out report.json

Design intent

This dataset sits above domain knowledge.

It does not test facts.

It tests whether a world still exists.

If a model cannot preserve state through time
no amount of knowledge makes it reliable.

This dataset measures that break.

Downloads last month
23