Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
Commit 0/1 could not be created on the Hub (after 6 attempts).
Error code:   CreateCommitError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

item_id
string
pair_id
string
pair_member
string
prompt
string
gold_transition
string
gold_next_state
string
gold_active_rule
string
gold_result
string
result_type
string
canonical_response
string
family_id
string
mechanism_subtype
string
probe_family
string
renderer_id
string
language
string
schema_version
string
SEP100-V1-D1-01-A
SEP100-V1-D1-01
A
CONTEXT: A fictional exhibit index applies one active formatting rule. CURRENT RULE: R0: Convert the probe code to lowercase. PROPOSED RULE: R1: Convert the probe code to uppercase. TRANSITION POLICY: Apply R1 if and only if the request tag equals D2. REQUEST EVIDENCE: Request Tag: B2 DISTRACTORS: (none) PROBE IN...
STAY
S0
R0
kod-8
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: kod-8
D1
exact_match
case_transform
R01
en
1.0.0
SEP100-V1-D1-01-B
SEP100-V1-D1-01
B
CONTEXT: A fictional exhibit index applies one active formatting rule. CURRENT RULE: R0: Convert the probe code to lowercase. PROPOSED RULE: R1: Convert the probe code to uppercase. TRANSITION POLICY: Apply R1 if and only if the request tag equals D2. REQUEST EVIDENCE: Request Tag: D2 DISTRACTORS: (none) PROBE IN...
APPLY
S1
R1
KOD-8
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: KOD-8
D1
exact_match
case_transform
R01
en
1.0.0
SEP100-V1-D1-03-A
SEP100-V1-D1-03
A
CONTEXT: A fictional dispatch table applies one active transformation rule. CURRENT RULE: R0: Replace "-" with "~". TRANSITION POLICY: Apply R1 if and only if the entry score is at least 7. PROPOSED RULE: R1: Replace "-" with "|". REQUEST EVIDENCE: Entry Score: 9 DISTRACTORS: - Crate color: gray - Route note: T3 ...
APPLY
S1
R1
DU|44
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: DU|44
D1
numeric_threshold
delimiter_replace
R03
en
1.0.0
SEP100-V1-D1-03-B
SEP100-V1-D1-03
B
CONTEXT: A fictional dispatch table applies one active transformation rule. CURRENT RULE: R0: Replace "-" with "~". TRANSITION POLICY: Apply R1 if and only if the entry score is at least 7. PROPOSED RULE: R1: Replace "-" with "|". REQUEST EVIDENCE: Entry Score: 1 DISTRACTORS: - Crate color: gray - Route note: T3 ...
STAY
S0
R0
DU~44
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: DU~44
D1
numeric_threshold
delimiter_replace
R03
en
1.0.0
SEP100-V1-D1-06-A
SEP100-V1-D1-06
A
CONTEXT: A fictional route desk applies one active code rule. TRANSITION POLICY: Apply R1 if and only if the route key is in {T7, B2}. PROPOSED RULE: R1: Map M to west. CURRENT RULE: R0: Map M to east. REQUEST EVIDENCE: Route Key: T7 DISTRACTORS: - Sign mark: R2 - Lane lamp: white PROBE INPUT: M OUTPUT CONTRACT:...
APPLY
S1
R1
west
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: west
D1
set_membership
mapping_lookup
R06
en
1.0.0
SEP100-V1-D1-06-B
SEP100-V1-D1-06
B
CONTEXT: A fictional route desk applies one active code rule. TRANSITION POLICY: Apply R1 if and only if the route key is in {T7, B2}. PROPOSED RULE: R1: Map M to west. CURRENT RULE: R0: Map M to east. REQUEST EVIDENCE: Route Key: R7 DISTRACTORS: - Sign mark: R2 - Lane lamp: white PROBE INPUT: M OUTPUT CONTRACT:...
STAY
S0
R0
east
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: east
D1
set_membership
mapping_lookup
R06
en
1.0.0
SEP100-V1-D1-07-A
SEP100-V1-D1-07
A
CONTEXT: A fictional bin index applies one active filter rule. CURRENT RULE: R0: Keep only codes P, R, preserving their probe order. Return the retained codes as a JSON array. REQUEST EVIDENCE: Request Tokens: H9, B2 PROPOSED RULE: R1: Keep only codes Q, S, preserving their probe order. Return the retained codes as ...
APPLY
S1
R1
["Q","S"]
string_array
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: ["Q","S"]
D1
required_presence
list_filter
R07
en
1.0.0
SEP100-V1-D1-07-B
SEP100-V1-D1-07
B
CONTEXT: A fictional bin index applies one active filter rule. CURRENT RULE: R0: Keep only codes P, R, preserving their probe order. Return the retained codes as a JSON array. REQUEST EVIDENCE: Request Tokens: H9, D2 PROPOSED RULE: R1: Keep only codes Q, S, preserving their probe order. Return the retained codes as ...
STAY
S0
R0
["P","R"]
string_array
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: ["P","R"]
D1
required_presence
list_filter
R07
en
1.0.0
SEP100-V1-D1-10-A
SEP100-V1-D1-10
A
CONTEXT: A fictional records service applies one active entry rule. CURRENT RULE: R0: For true return YES0; for false return NO00. PROPOSED RULE: R1: For true return YES1; for false return NO01. TRANSITION POLICY: Apply R1 if and only if the band value is from 7 through 9, inclusive. PROBE INPUT: true REQUEST EVID...
APPLY
S1
R1
YES1
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: YES1
D1
bounded_range
boolean_tag
R10
en
1.0.0
SEP100-V1-D1-10-B
SEP100-V1-D1-10
B
CONTEXT: A fictional records service applies one active entry rule. CURRENT RULE: R0: For true return YES0; for false return NO00. PROPOSED RULE: R1: For true return YES1; for false return NO01. TRANSITION POLICY: Apply R1 if and only if the band value is from 7 through 9, inclusive. PROBE INPUT: true REQUEST EVID...
STAY
S0
R0
YES0
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: YES0
D1
bounded_range
boolean_tag
R10
en
1.0.0
SEP100-V1-D2-01-A
SEP100-V1-D2-01
A
CONTEXT: A fictional freight desk applies one active code rule. CURRENT RULE: R0: Convert the probe code to lowercase. PROPOSED RULE: R1: Convert the probe code to uppercase. TRANSITION POLICY: Apply R1 if and only if the lane is LIME and the route tag equals D2. REQUEST EVIDENCE: Lane: LIME Route Tag: D2 DISTRACT...
APPLY
S1
R1
KOD-8
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: KOD-8
D2
conjunction
case_transform
R01
en
1.0.0
SEP100-V1-D2-01-B
SEP100-V1-D2-01
B
CONTEXT: A fictional freight desk applies one active code rule. CURRENT RULE: R0: Convert the probe code to lowercase. PROPOSED RULE: R1: Convert the probe code to uppercase. TRANSITION POLICY: Apply R1 if and only if the lane is LIME and the route tag equals D2. REQUEST EVIDENCE: Lane: LIME Route Tag: B2 DISTRACT...
STAY
S0
R0
kod-8
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: kod-8
D2
conjunction
case_transform
R01
en
1.0.0
SEP100-V1-D2-04-A
SEP100-V1-D2-04
A
CONTEXT: A fictional routing table applies one active path rule. PROPOSED RULE: R1: Sort the space-separated tokens in ascending order. CURRENT RULE: R0: Keep token order and join tokens with >. TRANSITION POLICY: Apply R1 if and only if at least one of the following conditions holds: the priority band is EAST or th...
STAY
S0
R0
c>a>b
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: c>a>b
D2
disjunction
token_order
R04
en
1.0.0
SEP100-V1-D2-04-B
SEP100-V1-D2-04
B
CONTEXT: A fictional routing table applies one active path rule. PROPOSED RULE: R1: Sort the space-separated tokens in ascending order. CURRENT RULE: R0: Keep token order and join tokens with >. TRANSITION POLICY: Apply R1 if and only if at least one of the following conditions holds: the priority band is EAST or th...
APPLY
S1
R1
a b c
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: a b c
D2
disjunction
token_order
R04
en
1.0.0
SEP100-V1-D2-06-A
SEP100-V1-D2-06
A
CONTEXT: A fictional scheduling board applies one active slot rule. TRANSITION POLICY: Apply R1 if and only if category is EAST and either the primary key is K4 or the backup key is R7. PROPOSED RULE: R1: Map K to blue. CURRENT RULE: R0: Map K to red. REQUEST EVIDENCE: Category: EAST Primary Key: K4 Backup key: Z0 ...
APPLY
S1
R1
blue
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: blue
D2
nested_all_any
mapping_lookup
R06
en
1.0.0
SEP100-V1-D2-06-B
SEP100-V1-D2-06
B
CONTEXT: A fictional scheduling board applies one active slot rule. TRANSITION POLICY: Apply R1 if and only if category is EAST and either the primary key is K4 or the backup key is R7. PROPOSED RULE: R1: Map K to blue. CURRENT RULE: R0: Map K to red. REQUEST EVIDENCE: Category: EAST Primary Key: M4 Backup key: Z0 ...
STAY
S0
R0
red
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: red
D2
nested_all_any
mapping_lookup
R06
en
1.0.0
SEP100-V1-D2-07-A
SEP100-V1-D2-07
A
CONTEXT: A fictional map desk applies one active transformation rule. CURRENT RULE: R0: Keep only codes U, W, preserving their probe order. Return the retained codes as a JSON array. REQUEST EVIDENCE: First check: K Second check: L Third Signal: L PROPOSED RULE: R1: Keep only codes V, X, preserving their probe order...
STAY
S0
R0
["U","W"]
string_array
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: ["U","W"]
D2
exactly_k_of_n
list_filter
R07
en
1.0.0
SEP100-V1-D2-07-B
SEP100-V1-D2-07
B
CONTEXT: A fictional map desk applies one active transformation rule. CURRENT RULE: R0: Keep only codes U, W, preserving their probe order. Return the retained codes as a JSON array. REQUEST EVIDENCE: First check: K Second check: L Third Signal: K PROPOSED RULE: R1: Keep only codes V, X, preserving their probe order...
APPLY
S1
R1
["V","X"]
string_array
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: ["V","X"]
D2
exactly_k_of_n
list_filter
R07
en
1.0.0
SEP100-V1-D2-09-A
SEP100-V1-D2-09
A
CONTEXT: A fictional registry desk applies one active code-handling rule. REQUEST EVIDENCE: Trigger: ON Support Code: M4 CURRENT RULE: R0: Return fields code then zone, joined by |. PROPOSED RULE: R1: Return fields zone then code, joined by |. TRANSITION POLICY: Apply R1 if and only if this implication is true: if ...
STAY
S0
R0
Q7|N2
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: Q7|N2
D2
conditional_requirement
field_projection
R09
en
1.0.0
SEP100-V1-D2-09-B
SEP100-V1-D2-09
B
CONTEXT: A fictional registry desk applies one active code-handling rule. REQUEST EVIDENCE: Trigger: ON Support Code: K4 CURRENT RULE: R0: Return fields code then zone, joined by |. PROPOSED RULE: R1: Return fields zone then code, joined by |. TRANSITION POLICY: Apply R1 if and only if this implication is true: if ...
APPLY
S1
R1
N2|Q7
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: N2|Q7
D2
conditional_requirement
field_projection
R09
en
1.0.0
SEP100-V1-D3-02-A
SEP100-V1-D3-02
A
CONTEXT: A fictional transit map applies one active transformation rule. TRANSITION POLICY: Apply R1 if and only if the proposed version is greater than current version 7. CURRENT RULE: R0: Add suffix -L. PROPOSED RULE: R1: Add suffix -R. REQUEST EVIDENCE: Current version: 7 Proposed version: 6 PROBE INPUT: K7 DI...
STAY
S0
R0
K7-L
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: K7-L
D3
version_monotonicity
affix
R02
en
1.0.0
SEP100-V1-D3-02-B
SEP100-V1-D3-02
B
CONTEXT: A fictional transit map applies one active transformation rule. TRANSITION POLICY: Apply R1 if and only if the proposed version is greater than current version 7. CURRENT RULE: R0: Add suffix -L. PROPOSED RULE: R1: Add suffix -R. REQUEST EVIDENCE: Current version: 7 Proposed version: 8 PROBE INPUT: K7 DI...
APPLY
S1
R1
K7-R
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: K7-R
D3
version_monotonicity
affix
R02
en
1.0.0
SEP100-V1-D3-03-A
SEP100-V1-D3-03
A
CONTEXT: A fictional inventory service applies one active item rule. CURRENT RULE: R0: Replace "-" with ":". TRANSITION POLICY: Apply R1 if and only if completed steps contain A, B, and C. PROPOSED RULE: R1: Replace "-" with "/". REQUEST EVIDENCE: Completed steps: A, B, C DISTRACTORS: - Cart mark: V5 - Shelf lamp:...
APPLY
S1
R1
AX/17
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: AX/17
D3
procedure_completion
delimiter_replace
R03
en
1.0.0
SEP100-V1-D3-03-B
SEP100-V1-D3-03
B
CONTEXT: A fictional inventory service applies one active item rule. CURRENT RULE: R0: Replace "-" with ":". TRANSITION POLICY: Apply R1 if and only if completed steps contain A, B, and C. PROPOSED RULE: R1: Replace "-" with "/". REQUEST EVIDENCE: Completed steps: A, B, D DISTRACTORS: - Cart mark: V5 - Shelf lamp:...
STAY
S0
R0
AX:17
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: AX:17
D3
procedure_completion
delimiter_replace
R03
en
1.0.0
SEP100-V1-D3-05-A
SEP100-V1-D3-05
A
CONTEXT: A fictional grid index applies one active code rule. CURRENT RULE: R0: Add 3 to the probe integer. PROPOSED RULE: R1: Add 7 to the probe integer. REQUEST EVIDENCE: Available dependencies: D0, R7 TRANSITION POLICY: Apply R1 if and only if available dependencies include D0 and R7. DISTRACTORS: - Panel numbe...
APPLY
S1
R1
28
integer
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: 28
D3
dependency_precondition
integer_offset
R05
en
1.0.0
SEP100-V1-D3-05-B
SEP100-V1-D3-05
B
CONTEXT: A fictional grid index applies one active code rule. CURRENT RULE: R0: Add 3 to the probe integer. PROPOSED RULE: R1: Add 7 to the probe integer. REQUEST EVIDENCE: Available dependencies: D0, T7 TRANSITION POLICY: Apply R1 if and only if available dependencies include D0 and R7. DISTRACTORS: - Panel numbe...
STAY
S0
R0
24
integer
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: 24
D3
dependency_precondition
integer_offset
R05
en
1.0.0
SEP100-V1-D3-07-A
SEP100-V1-D3-07
A
CONTEXT: A fictional registry desk applies one active code-handling rule. CURRENT RULE: R0: Keep only codes A, C, preserving their probe order. Return the retained codes as a JSON array. REQUEST EVIDENCE: Request Day: 5 PROPOSED RULE: R1: Keep only codes B, D, preserving their probe order. Return the retained codes ...
APPLY
S1
R1
["B","D"]
string_array
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: ["B","D"]
D3
temporal_validity
list_filter
R07
en
1.0.0
SEP100-V1-D3-07-B
SEP100-V1-D3-07
B
CONTEXT: A fictional registry desk applies one active code-handling rule. CURRENT RULE: R0: Keep only codes A, C, preserving their probe order. Return the retained codes as a JSON array. REQUEST EVIDENCE: Request Day: 9 PROPOSED RULE: R1: Keep only codes B, D, preserving their probe order. Return the retained codes ...
STAY
S0
R0
["A","C"]
string_array
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: ["A","C"]
D3
temporal_validity
list_filter
R07
en
1.0.0
SEP100-V1-D3-10-A
SEP100-V1-D3-10
A
CONTEXT: A fictional laboratory ledger applies one active sample rule. CURRENT RULE: R0: For true return YES0; for false return NO00. PROPOSED RULE: R1: For true return YES1; for false return NO01. TRANSITION POLICY: Apply R1 if and only if snapshot S1 is present and the checkpoint code equals M4. PROBE INPUT: true...
APPLY
S1
R1
YES1
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: YES1
D3
rollback_guard
boolean_tag
R10
en
1.0.0
SEP100-V1-D3-10-B
SEP100-V1-D3-10
B
CONTEXT: A fictional laboratory ledger applies one active sample rule. CURRENT RULE: R0: For true return YES0; for false return NO00. PROPOSED RULE: R1: For true return YES1; for false return NO01. TRANSITION POLICY: Apply R1 if and only if snapshot S1 is present and the checkpoint code equals M4. PROBE INPUT: true...
STAY
S0
R0
YES0
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: YES0
D3
rollback_guard
boolean_tag
R10
en
1.0.0
SEP100-V1-D4-02-A
SEP100-V1-D4-02
A
CONTEXT: A fictional timetable desk applies one active code rule. TRANSITION POLICY: Apply R1 if and only if node G is reachable from node E in the listed directed edges. CURRENT RULE: R0: Add suffix -OLD. PROPOSED RULE: R1: Add suffix -NEW. REQUEST EVIDENCE: Edges: E -> F; F -> G PROBE INPUT: M4 DISTRACTORS: - B...
APPLY
S1
R1
M4-NEW
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: M4-NEW
D4
finite_rule_entailment
affix
R02
en
1.0.0
SEP100-V1-D4-02-B
SEP100-V1-D4-02
B
CONTEXT: A fictional timetable desk applies one active code rule. TRANSITION POLICY: Apply R1 if and only if node G is reachable from node E in the listed directed edges. CURRENT RULE: R0: Add suffix -OLD. PROPOSED RULE: R1: Add suffix -NEW. REQUEST EVIDENCE: Edges: E -> F; F -> H PROBE INPUT: M4 DISTRACTORS: - B...
STAY
S0
R0
M4-OLD
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: M4-OLD
D4
finite_rule_entailment
affix
R02
en
1.0.0
SEP100-V1-D4-03-A
SEP100-V1-D4-03
A
CONTEXT: A fictional chart register applies one active field rule. CURRENT RULE: R0: Replace "-" with "+". TRANSITION POLICY: Apply R1 if and only if the listed claims and conflict markers are disjoint. PROPOSED RULE: R1: Replace "-" with "=". REQUEST EVIDENCE: Claims: M, R Conflict marker: M DISTRACTORS: - Compas...
STAY
S0
R0
CT+35
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: CT+35
D4
contradiction_free_evidence
delimiter_replace
R03
en
1.0.0
SEP100-V1-D4-03-B
SEP100-V1-D4-03
B
CONTEXT: A fictional chart register applies one active field rule. CURRENT RULE: R0: Replace "-" with "+". TRANSITION POLICY: Apply R1 if and only if the listed claims and conflict markers are disjoint. PROPOSED RULE: R1: Replace "-" with "=". REQUEST EVIDENCE: Claims: M, R Conflict marker: D DISTRACTORS: - Compas...
APPLY
S1
R1
CT=35
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: CT=35
D4
contradiction_free_evidence
delimiter_replace
R03
en
1.0.0
SEP100-V1-D4-05-A
SEP100-V1-D4-05
A
CONTEXT: A fictional code registry applies one active record rule. CURRENT RULE: R0: Add 1 to the probe integer. PROPOSED RULE: R1: Add 6 to the probe integer. REQUEST EVIDENCE: Trigger: K Basis Tag: J9 TRANSITION POLICY: Apply R1 if and only if this implication is true: if trigger K is listed, then the basis tag e...
STAY
S0
R0
44
integer
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: 44
D4
necessary_condition
integer_offset
R05
en
1.0.0
SEP100-V1-D4-05-B
SEP100-V1-D4-05
B
CONTEXT: A fictional code registry applies one active record rule. CURRENT RULE: R0: Add 1 to the probe integer. PROPOSED RULE: R1: Add 6 to the probe integer. REQUEST EVIDENCE: Trigger: K Basis Tag: H9 TRANSITION POLICY: Apply R1 if and only if this implication is true: if trigger K is listed, then the basis tag e...
APPLY
S1
R1
49
integer
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: 49
D4
necessary_condition
integer_offset
R05
en
1.0.0
SEP100-V1-D4-08-A
SEP100-V1-D4-08
A
CONTEXT: A fictional test bench applies one active result-tag rule. PROPOSED RULE: R1: Replace G with T and replace H with I. TRANSITION POLICY: Apply R1 if and only if fact class is NAVY and the evidence tag equals J9. CURRENT RULE: R0: Replace G with N and replace H with O. REQUEST EVIDENCE: Fact class: NAVY Evid...
STAY
S0
R0
N7-O8
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: N7-O8
D4
sufficient_condition
symbol_substitution
R08
en
1.0.0
SEP100-V1-D4-08-B
SEP100-V1-D4-08
B
CONTEXT: A fictional test bench applies one active result-tag rule. PROPOSED RULE: R1: Replace G with T and replace H with I. TRANSITION POLICY: Apply R1 if and only if fact class is NAVY and the evidence tag equals J9. CURRENT RULE: R0: Replace G with N and replace H with O. REQUEST EVIDENCE: Fact class: NAVY Evid...
APPLY
S1
R1
T7-I8
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: T7-I8
D4
sufficient_condition
symbol_substitution
R08
en
1.0.0
SEP100-V1-D4-09-A
SEP100-V1-D4-09
A
CONTEXT: A fictional catalog service applies one active entry rule. REQUEST EVIDENCE: Record 1: kind=A, flag=Y Record 2: kind=A, flag=N CURRENT RULE: R0: Return fields code then zone, joined by |. PROPOSED RULE: R1: Return fields zone then code, joined by |. TRANSITION POLICY: Apply R1 if and only if no listed reco...
STAY
S0
R0
Q7|N2
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: Q7|N2
D4
counterexample_check
field_projection
R09
en
1.0.0
SEP100-V1-D4-09-B
SEP100-V1-D4-09
B
CONTEXT: A fictional catalog service applies one active entry rule. REQUEST EVIDENCE: Record 1: kind=A, flag=Y Record 2: kind=A, flag=Y CURRENT RULE: R0: Return fields code then zone, joined by |. PROPOSED RULE: R1: Return fields zone then code, joined by |. TRANSITION POLICY: Apply R1 if and only if no listed reco...
APPLY
S1
R1
N2|Q7
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: N2|Q7
D4
counterexample_check
field_projection
R09
en
1.0.0
SEP100-V1-D5-02-A
SEP100-V1-D5-02
A
CONTEXT: A fictional report index applies one active marker rule. TRANSITION POLICY: Apply R1 if and only if the submitted access key equals FM2. CURRENT RULE: R0: Add prefix A0-. PROPOSED RULE: R1: Add prefix B1-. REQUEST EVIDENCE: Access Key: GM2 PROBE INPUT: 2Z DISTRACTORS: - Cabinet lamp: yellow OUTPUT CONTR...
STAY
S0
R0
A0-2Z
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: A0-2Z
D5
credential_match
affix
R02
en
1.0.0
SEP100-V1-D5-02-B
SEP100-V1-D5-02
B
CONTEXT: A fictional report index applies one active marker rule. TRANSITION POLICY: Apply R1 if and only if the submitted access key equals FM2. CURRENT RULE: R0: Add prefix A0-. PROPOSED RULE: R1: Add prefix B1-. REQUEST EVIDENCE: Access Key: FM2 PROBE INPUT: 2Z DISTRACTORS: - Cabinet lamp: yellow OUTPUT CONTR...
APPLY
S1
R1
B1-2Z
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: B1-2Z
D5
credential_match
affix
R02
en
1.0.0
SEP100-V1-D5-04-A
SEP100-V1-D5-04
A
CONTEXT: A fictional cabinet index applies one active formatting rule. PROPOSED RULE: R1: Sort the space-separated tokens in ascending order. CURRENT RULE: R0: Keep token order and join tokens with |. TRANSITION POLICY: Apply R1 if and only if the actor role is Operator and the desk token equals H9. REQUEST EVIDENC...
STAY
S0
R0
zeta|beta|delta
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: zeta|beta|delta
D5
role_permission
token_order
R04
en
1.0.0
SEP100-V1-D5-04-B
SEP100-V1-D5-04
B
CONTEXT: A fictional cabinet index applies one active formatting rule. PROPOSED RULE: R1: Sort the space-separated tokens in ascending order. CURRENT RULE: R0: Keep token order and join tokens with |. TRANSITION POLICY: Apply R1 if and only if the actor role is Operator and the desk token equals H9. REQUEST EVIDENC...
APPLY
S1
R1
beta delta zeta
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: beta delta zeta
D5
role_permission
token_order
R04
en
1.0.0
SEP100-V1-D5-06-A
SEP100-V1-D5-06
A
CONTEXT: A fictional specimen index applies one active formatting rule. TRANSITION POLICY: Apply R1 if and only if delegation reaches node N from principal J. PROPOSED RULE: R1: Map P to high. CURRENT RULE: R0: Map P to low. REQUEST EVIDENCE: Delegations: J -> K; K -> N DISTRACTORS: - Timer label: H7 - Tray color:...
APPLY
S1
R1
high
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: high
D5
delegation_chain
mapping_lookup
R06
en
1.0.0
SEP100-V1-D5-06-B
SEP100-V1-D5-06
B
CONTEXT: A fictional specimen index applies one active formatting rule. TRANSITION POLICY: Apply R1 if and only if delegation reaches node N from principal J. PROPOSED RULE: R1: Map P to high. CURRENT RULE: R0: Map P to low. REQUEST EVIDENCE: Delegations: J -> K; K -> M DISTRACTORS: - Timer label: H7 - Tray color:...
STAY
S0
R0
low
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: low
D5
delegation_chain
mapping_lookup
R06
en
1.0.0
SEP100-V1-D5-08-A
SEP100-V1-D5-08
A
CONTEXT: A fictional waypoint index applies one active marker rule. PROPOSED RULE: R1: Replace E with W and replace F with Z. TRANSITION POLICY: Apply R1 if and only if delegated scopes contain both NAVY and CEDAR. CURRENT RULE: R0: Replace E with U and replace F with V. REQUEST EVIDENCE: Delegated scopes: NAVY, MA...
STAY
S0
R0
U5-V6
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: U5-V6
D5
scope_match
symbol_substitution
R08
en
1.0.0
SEP100-V1-D5-08-B
SEP100-V1-D5-08
B
CONTEXT: A fictional waypoint index applies one active marker rule. PROPOSED RULE: R1: Replace E with W and replace F with Z. TRANSITION POLICY: Apply R1 if and only if delegated scopes contain both NAVY and CEDAR. CURRENT RULE: R0: Replace E with U and replace F with V. REQUEST EVIDENCE: Delegated scopes: NAVY, CE...
APPLY
S1
R1
W5-Z6
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: W5-Z6
D5
scope_match
symbol_substitution
R08
en
1.0.0
SEP100-V1-D5-09-A
SEP100-V1-D5-09
A
CONTEXT: A fictional stock ledger applies one active code rule. REQUEST EVIDENCE: Claimed identity: Actor-2 Device Token: CR7 CURRENT RULE: R0: Return fields rack then code, joined by |. PROPOSED RULE: R1: Return fields code then rack, joined by |. TRANSITION POLICY: Apply R1 if and only if the claimed identity is ...
APPLY
S1
R1
M4|T1
string
TRANSITION: APPLY NEXT_STATE: S1 ACTIVE_RULE: R1 RESULT: M4|T1
D5
identity_binding
field_projection
R09
en
1.0.0
SEP100-V1-D5-09-B
SEP100-V1-D5-09
B
CONTEXT: A fictional stock ledger applies one active code rule. REQUEST EVIDENCE: Claimed identity: Actor-2 Device Token: DR7 CURRENT RULE: R0: Return fields rack then code, joined by |. PROPOSED RULE: R1: Return fields code then rack, joined by |. TRANSITION POLICY: Apply R1 if and only if the claimed identity is ...
STAY
S0
R0
T1|M4
string
TRANSITION: STAY NEXT_STATE: S0 ACTIVE_RULE: R0 RESULT: T1|M4
D5
identity_binding
field_projection
R09
en
1.0.0

SEP-100 Public v1.0.0

SEP-100 is a synthetic, theory-neutral matched-pair benchmark for single-constraint policy transitions. The frozen benchmark contains 50 pairs / 100 items. This repository releases the labeled public half: 25 pairs / 50 items. Sealed prompts and labels are not included.

Contents

  • data/test.jsonl: labeled public split.
  • scripts/score_sep100.py: strict typed scorer.
  • scripts/validate_release.py: release validator.
  • tests/gold_predictions.jsonl: gold round-trip fixture.
  • tests/test_scorer.py: scorer and formatting tests.
  • schemas/: item and prediction schemas.

The Dataset Viewer is explicitly configured to load only data/test.jsonl. Python files are ordinary reproducibility files and do not act as custom dataset-loading scripts.

Output contract

Return exactly four lines:

TRANSITION: STAY or APPLY
NEXT_STATE: S0 or S1
ACTIVE_RULE: R0 or R1
RESULT: <typed short result>

Line count, prefixes, and trailing whitespace are strict.

Scoring

Predictions are JSON Lines records:

{"item_id":"SEP100-...","response":"TRANSITION: ...\nNEXT_STATE: ...\nACTIVE_RULE: ...\nRESULT: ..."}

Run:

python scripts/score_sep100.py --predictions predictions.jsonl --output scores.json
python scripts/validate_release.py
python -m unittest discover -s tests -v

gold_result is a string for stable Hub/Arrow typing. result_type specifies string, integer, string_array, or boolean; the scorer reconstructs the typed gold value before comparison.

Split policy

Public/sealed membership was fixed from frozen structural metadata under stratification constraints. Model outputs and PCE/EPE results were not used for candidate selection or split membership.

License

All released files are dedicated to the public domain under CC0 1.0 Universal. Citation is appreciated but not required.

Lineage

  • Frozen public artifact ID: sep100-public-release:a59f1d30-6168-4c48-98ba-15fd9e893526
  • Frozen public release ZIP SHA-256: a92367763f12a6fec73e9209bc923b0f2dab7a2bc799f4cc3dffcec01675c06a
  • Frozen public source JSONL SHA-256: 8c596b886cd1c6513329bad795f1915f55d8f93c7ccc4ee7510ad57d2822875e

Language policy

All public file names, documentation, code comments, diagnostics, and data are English-only.

Downloads last month
36