axondendriteplus/decipop-v7-r2-control
Text Classification • 0.4B • Updated • 17
Datasets:
example_id stringlengths 45 61 | family_id stringlengths 28 32 | split stringclasses 1
value | theory dict | query dict | label stringclasses 3
values | proof_depth int64 1 3 ⌀ | theory_depth int64 0 3 | motif stringclasses 4
values | renderer_id stringclasses 2
values | theory_hash stringlengths 64 64 | structure_hash stringclasses 14
values | facts listlengths 6 15 | rules listlengths 2 4 | question stringlengths 13 26 | prompt stringlengths 578 1.01k |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
v7-r1-calibration-old-000085-calibration-false | v7-r1-calibration-old-000085 | calibration | {
"entities": [
"Evan",
"EntityA",
"Gabe",
"Farah"
],
"facts": [
{
"predicate": "bright",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "calm",
"arguments": [
"Evan"
],
"negative": false
},
{
... | {
"predicate": "calm",
"arguments": [
"Gabe"
],
"negative": false
} | FALSE | 3 | 3 | relational | canonical | ac10499750de1680de85c32d8e684d3fdcd50499dafc7dc73fbd4a20d7b99675 | ca1f529f2d76e8e33e8e05b33259f6d92c1740d6605cd6d61b5f2c5d1ef0eb0c | [
"EntityA supports Evan.",
"Farah is calm.",
"Gabe is not round.",
"Evan is quiet.",
"EntityA is bright.",
"Evan is not kind.",
"Evan is calm.",
"Gabe knows EntityA."
] | [
"For every person mentioned: if person 1 supports person 2 and person 2 is not kind, then person 1 is not large.",
"For every person mentioned: if person 1 knows person 2 and person 2 is not large, then person 1 is quiet.",
"For every person mentioned: if person 1 is quiet and person 1 is round, then person 1 i... | Gabe is calm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
EntityA supports Evan.
Farah is calm.
Gabe is not round.
Evan is quiet.
EntityA is brigh... |
v7-r1-calibration-old-000118-calibration-unknown | v7-r1-calibration-old-000118 | calibration | {
"entities": [
"Dalia",
"EntityA",
"Chen",
"Gabe"
],
"facts": [
{
"predicate": "bright",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": "calm",
"arguments": [
"Chen"
],
"negative": true
},
{
"... | {
"predicate": "blue",
"arguments": [
"EntityA"
],
"negative": true
} | UNKNOWN | null | 1 | relational | canonical | a2fd26d79889fd6f66d8ec3620b8a86c6f6030e42134aea285e13ae7f1c13cf6 | 999d85ff01ce586cf52d04fd5e2d4008e5d57dcf3fdbd38c3b52abceb1cc69b3 | [
"Chen is kind.",
"Dalia is bright.",
"Gabe is not quiet.",
"Chen is not calm.",
"Chen is not warm.",
"EntityA is not kind.",
"Gabe is round.",
"Gabe is warm.",
"Dalia is not large."
] | [
"For every person mentioned: if person 1 is bright and person 1 is not large, then person 1 is blue.",
"For every person mentioned: if person 1 is bright and person 1 is large, then person 1 is not blue."
] | EntityA is not blue. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is kind.
Dalia is bright.
Gabe is not quiet.
Chen is not calm.
Chen is not warm.
En... |
v7-r1-calibration-old-000043-calibration-false | v7-r1-calibration-old-000043 | calibration | {
"entities": [
"Dalia",
"Hana",
"Chen"
],
"facts": [
{
"predicate": "bright",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "bright",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": ... | {
"predicate": "round",
"arguments": [
"Hana"
],
"negative": false
} | FALSE | 2 | 2 | relational | canonical | 1cb94326e94b41ffab31cc81edefc0c617974bbd327b36ecab60d987344688a1 | 44683fac69fbc06ccbf74f16e30207e66c2ec7d4a3a61270b430a85a65b794c8 | [
"Dalia is bright.",
"Chen is quiet.",
"Dalia is large.",
"Chen is bright.",
"Hana is warm.",
"Dalia is calm.",
"Hana follows Dalia."
] | [
"For every person mentioned: if person 1 is kind and person 1 is not warm, then person 1 is round.",
"For every person mentioned: if person 1 is kind and person 1 is warm, then person 1 is not round.",
"For every person mentioned: if person 1 follows person 2 and person 2 is large, then person 1 is kind."
] | Hana is round. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is bright.
Chen is quiet.
Dalia is large.
Chen is bright.
Hana is warm.
Dalia is c... |
v7-r1-calibration-old-000050-calibration-true | v7-r1-calibration-old-000050 | calibration | {
"entities": [
"Chen",
"Dalia",
"Hana"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "ki... | {
"predicate": "warm",
"arguments": [
"Chen"
],
"negative": false
} | TRUE | 3 | 3 | chain | conditional | a609bb4e706354d99f367d88f496954b28b729ce66a0a77950d001cdc5c9f98b | 8628c5aad256c8edba6f7f2657c0483db19285b433a26e9bbaccd959071fc344 | [
"Chen is bright.",
"Hana is not round.",
"Dalia is not kind.",
"Chen is not large.",
"Dalia is not blue.",
"Dalia is quiet."
] | [
"Whenever person 1 is blue, person 1 is quiet; this applies to all people mentioned.",
"Whenever person 1 is not large, person 1 is blue; this applies to all people mentioned.",
"Whenever person 1 is quiet and person 1 is not bright, person 1 is not warm; this applies to all people mentioned.",
"Whenever pers... | Chen is warm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is bright.
Hana is not round.
Dalia is not kind.
Chen is not large.
Dalia is not bl... |
v7-r1-calibration-old-000023-calibration-true | v7-r1-calibration-old-000023 | calibration | {
"entities": [
"Hana",
"EntityA",
"Farah",
"Gabe"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Gabe"
],
"negative": true
},
{
"predicate": "authorized",
"arguments": [
"Farah"
],
"negative": true
},
{
... | {
"predicate": "trusted",
"arguments": [
"Hana"
],
"negative": true
} | TRUE | 3 | 3 | chain | canonical | 24c58f015eb0bd0ebf3e9363f5a0c47dd6303367b38508364d7befd0dcd3e5a7 | 8628c5aad256c8edba6f7f2657c0483db19285b433a26e9bbaccd959071fc344 | [
"Gabe is not active.",
"Hana is not enabled.",
"Gabe is registered.",
"Hana is not ready.",
"Farah is not authorized.",
"EntityA is verified."
] | [
"For every person mentioned: if person 1 is not enabled, then person 1 is active.",
"For every person mentioned: if person 1 is registered and person 1 is ready, then person 1 is trusted.",
"For every person mentioned: if person 1 is active, then person 1 is registered.",
"For every person mentioned: if perso... | Hana is not trusted. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is not active.
Hana is not enabled.
Gabe is registered.
Hana is not ready.
Farah is... |
v7-r1-calibration-multi-5-0003-0-multi-base-4 | v7-r1-calibration-multi-5-0003-0 | calibration | {
"entities": [
"Dalia",
"Evan",
"Farah",
"Gabe",
"Hana"
],
"facts": [
{
"predicate": "quiet",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": "round",
"arguments": [
"Gabe"
],
"negative": true
},
... | {
"predicate": "kind",
"arguments": [
"Hana"
],
"negative": true
} | UNKNOWN | null | 2 | signed_chain | canonical | 7f1a49395988b1578e09454f3286ac4e607918ee081c786a3e5032cb7db30e98 | cdfd8d92931b901caba0f41203af2efb0f3e650038aa30522d44024f1a911cb1 | [
"Gabe is quiet.",
"Farah is not round.",
"Farah is warm.",
"Farah is calm.",
"Gabe is not round.",
"Evan is warm.",
"Evan is calm.",
"Evan is quiet.",
"Hana is calm.",
"Dalia is warm.",
"Dalia is calm.",
"Hana is warm.",
"Dalia is quiet.",
"Hana is quiet.",
"Gabe is warm."
] | [
"For every person mentioned: if person 1 is round, then person 1 is blue.",
"For every person mentioned: if person 1 is not round, then person 1 is not blue.",
"For every person mentioned: if person 1 is blue, then person 1 is kind.",
"For every person mentioned: if person 1 is not blue, then person 1 is not ... | Hana is not kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is quiet.
Farah is not round.
Farah is warm.
Farah is calm.
Gabe is not round.
Evan... |
v7-r1-calibration-old-000113-calibration-true | v7-r1-calibration-old-000113 | calibration | {
"entities": [
"Dalia",
"Chen",
"Bella",
"Hana"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "bright",
"arguments": [
"Bella"
],
"negative": true
},
{
"pr... | {
"predicate": "bright",
"arguments": [
"Dalia"
],
"negative": false
} | TRUE | 1 | 1 | relational | canonical | 7d4c341288f0b5fe1de0db85763cec3c14edbc4e0cc425caf039368ca74be951 | 5b711a9c568ac2fd187eae882b1a35a74b5549b53da284a9e55be83fad8e4b23 | [
"Dalia is not large.",
"Chen is calm.",
"Chen is round.",
"Hana is quiet.",
"Chen is blue.",
"Dalia is kind.",
"Bella is not bright."
] | [
"For every person mentioned: if person 1 is kind and person 1 is large, then person 1 is not bright.",
"For every person mentioned: if person 1 is kind and person 1 is not large, then person 1 is bright."
] | Dalia is bright. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is not large.
Chen is calm.
Chen is round.
Hana is quiet.
Chen is blue.
Dalia is k... |
v7-r1-calibration-old-000124-calibration-false | v7-r1-calibration-old-000124 | calibration | {
"entities": [
"Dalia",
"Bella",
"Hana",
"EntityA"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Bella"
],
"negative": true
},
{
"predicate": "approved",
"arguments": [
"Dalia"
],
"negative": false
},
{
... | {
"predicate": "ready",
"arguments": [
"Dalia"
],
"negative": true
} | FALSE | 3 | 3 | chain | conditional | 17938eb3165c7da742c9d1397c7b75d5e65cc8f4a7ad460bb0dbe96babb9e5d2 | 8628c5aad256c8edba6f7f2657c0483db19285b433a26e9bbaccd959071fc344 | [
"Bella is not active.",
"Bella is not enabled.",
"Bella is not trusted.",
"Dalia is authorized.",
"Dalia is approved.",
"Bella is registered."
] | [
"Whenever person 1 is authorized, person 1 is enabled; this applies to all people mentioned.",
"Whenever person 1 is active and person 1 is not approved, person 1 is not ready; this applies to all people mentioned.",
"Whenever person 1 is enabled, person 1 is active; this applies to all people mentioned.",
"W... | Dalia is not ready. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is not active.
Bella is not enabled.
Bella is not trusted.
Dalia is authorized.
Da... |
v7-r1-calibration-multi-4-0002-0-multi-base-1 | v7-r1-calibration-multi-4-0002-0 | calibration | {
"entities": [
"Chen",
"Dalia",
"Evan",
"Farah"
],
"facts": [
{
"predicate": "trusted",
"arguments": [
"Evan"
],
"negative": false
},
{
"predicate": "trusted",
"arguments": [
"Dalia"
],
"negative": false
},
{
... | {
"predicate": "enabled",
"arguments": [
"Dalia"
],
"negative": false
} | TRUE | 1 | 1 | conjunction | canonical | 9092186ececa1d4835c030728d0a4d0e7303a067a84d823cbafcd1c53f550968 | cadd90deb012a4b9628b7ec27362942c3e58b1b395b20e891e38650ad17af718 | [
"Dalia is registered.",
"Evan is registered.",
"Evan is trusted.",
"Dalia is active.",
"Farah is active.",
"Chen is active.",
"Farah is trusted.",
"Dalia is trusted.",
"Chen is registered.",
"Farah is ready.",
"Evan is active.",
"Chen is not trusted."
] | [
"For every person mentioned: if person 1 is active and person 1 is trusted, then person 1 is enabled.",
"For every person mentioned: if person 1 is active and person 1 is not trusted, then person 1 is not enabled."
] | Dalia is enabled. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is registered.
Evan is registered.
Evan is trusted.
Dalia is active.
Farah is acti... |
v7-r1-calibration-multi-5-0004-1-multi-base-4 | v7-r1-calibration-multi-5-0004-1 | calibration | {
"entities": [
"Evan",
"Farah",
"Gabe",
"Hana",
"EntityA"
],
"facts": [
{
"predicate": "trusted",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "verified",
"arguments": [
"Hana"
],
"negative": false
... | {
"predicate": "ready",
"arguments": [
"EntityA"
],
"negative": false
} | TRUE | 1 | 1 | conjunction | canonical | bf9318183ebe4948e96a953a14674421d2c51dbc14a4820f1b406743804a8df0 | cadd90deb012a4b9628b7ec27362942c3e58b1b395b20e891e38650ad17af718 | [
"Farah is trusted.",
"Evan is trusted.",
"Gabe is trusted.",
"EntityA is trusted.",
"Evan is not enabled.",
"Hana is verified.",
"Gabe is verified.",
"EntityA is enabled.",
"EntityA is verified.",
"Evan is verified.",
"Farah is verified.",
"Hana is authorized.",
"Farah is not enabled.",
"H... | [
"For every person mentioned: if person 1 is verified and person 1 is enabled, then person 1 is ready.",
"For every person mentioned: if person 1 is verified and person 1 is not enabled, then person 1 is not ready."
] | EntityA is ready. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Farah is trusted.
Evan is trusted.
Gabe is trusted.
EntityA is trusted.
Evan is not enab... |
v7-r1-calibration-old-000073-calibration-true | v7-r1-calibration-old-000073 | calibration | {
"entities": [
"Bella",
"EntityA",
"Dalia"
],
"facts": [
{
"predicate": "active",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "active",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predi... | {
"predicate": "trusted",
"arguments": [
"Bella"
],
"negative": false
} | TRUE | 1 | 1 | chain | conditional | a31b40d19c3c856dbc8ec27a8034b70cf2b87cdf8a793c3b2612f05938bdc032 | d6cdea0617b6466794837beafb7653bbf07acb9494a5c628e1b8f13f481fdb99 | [
"Bella is not authorized.",
"Bella is not verified.",
"EntityA is active.",
"Dalia is active.",
"EntityA is approved.",
"Dalia is registered.",
"Dalia is approved.",
"Dalia is ready.",
"Dalia is not enabled.",
"EntityA is not trusted."
] | [
"Whenever person 1 is not authorized and person 1 is not verified, person 1 is trusted; this applies to all people mentioned.",
"Whenever person 1 is not authorized and person 1 is verified, person 1 is not trusted; this applies to all people mentioned."
] | Bella is trusted. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is not authorized.
Bella is not verified.
EntityA is active.
Dalia is active.
Enti... |
v7-r1-calibration-old-000082-calibration-true | v7-r1-calibration-old-000082 | calibration | {
"entities": [
"Dalia",
"Bella",
"Gabe"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Bella"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": "... | {
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": true
} | TRUE | 3 | 3 | chain | canonical | 9ddbeb4377142b905513fab7c4d096dc365d6fc072152a7776f0429f9784895f | 8628c5aad256c8edba6f7f2657c0483db19285b433a26e9bbaccd959071fc344 | [
"Gabe is not kind.",
"Bella is not calm.",
"Bella is warm.",
"Bella is not blue.",
"Bella is large.",
"Bella is not kind.",
"Dalia is round.",
"Dalia is bright."
] | [
"For every person mentioned: if person 1 is not quiet and person 1 is not round, then person 1 is blue.",
"For every person mentioned: if person 1 is not quiet and person 1 is round, then person 1 is not blue.",
"For every person mentioned: if person 1 is bright, then person 1 is not kind.",
"For every person... | Dalia is not blue. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is not kind.
Bella is not calm.
Bella is warm.
Bella is not blue.
Bella is large.
B... |
v7-r1-calibration-old-000061-calibration-true | v7-r1-calibration-old-000061 | calibration | {
"entities": [
"Bella",
"Dalia",
"EntityA"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Bella"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate"... | {
"predicate": "kind",
"arguments": [
"Bella"
],
"negative": true
} | TRUE | 3 | 3 | chain | conditional | 9c6a927ec0ff74f200698a4c0c5d3839121174768b580630d1bd70cc5deaa635 | 8628c5aad256c8edba6f7f2657c0483db19285b433a26e9bbaccd959071fc344 | [
"EntityA is warm.",
"Bella is not blue.",
"EntityA is not calm.",
"Dalia is not kind.",
"Dalia is not quiet.",
"Dalia is bright.",
"Bella is not round."
] | [
"Whenever person 1 is not bright, person 1 is calm; this applies to all people mentioned.",
"Whenever person 1 is calm and person 1 is not round, person 1 is not kind; this applies to all people mentioned.",
"Whenever person 1 is calm and person 1 is round, person 1 is kind; this applies to all people mentioned... | Bella is not kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
EntityA is warm.
Bella is not blue.
EntityA is not calm.
Dalia is not kind.
Dalia is not... |
v7-r1-calibration-multi-4-0006-2-multi-support_edit-3 | v7-r1-calibration-multi-4-0006-2 | calibration | {
"entities": [
"Gabe",
"Hana",
"EntityA",
"Bella"
],
"facts": [
{
"predicate": "kind",
"arguments": [
"Bella"
],
"negative": false
},
{
"predicate": "round",
"arguments": [
"Gabe"
],
"negative": false
},
{
"... | {
"predicate": "quiet",
"arguments": [
"Bella"
],
"negative": false
} | UNKNOWN | null | 1 | conjunction | canonical | 4c7c1548c4310c2e71cc7ba5237a2a81cefcab7c7e3291f28fd29ac0971e9018 | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Bella is round.",
"Gabe is kind.",
"Hana is round.",
"Hana is kind.",
"EntityA is round.",
"Bella is bright.",
"EntityA is bright.",
"Gabe is bright.",
"Gabe is round.",
"EntityA is not calm.",
"Hana is bright.",
"Bella is kind."
] | [
"For every person mentioned: if person 1 is round and person 1 is not calm, then person 1 is not quiet.",
"For every person mentioned: if person 1 is round and person 1 is calm, then person 1 is quiet."
] | Bella is quiet. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is round.
Gabe is kind.
Hana is round.
Hana is kind.
EntityA is round.
Bella is br... |
v7-r1-calibration-old-000067-calibration-false | v7-r1-calibration-old-000067 | calibration | {
"entities": [
"Gabe",
"Farah",
"Evan",
"Chen"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "ready",
"arguments": [
"Farah"
],
"negative": false
},
{
"p... | {
"predicate": "ready",
"arguments": [
"Gabe"
],
"negative": true
} | FALSE | 2 | 2 | chain | conditional | e28fd408c09aef67f99159207648af06f18328f80f6100e72d13877321a92a32 | 90ced93ad4f5e7b3ebf82a3a74e62256c648008278ba895982d502605377148d | [
"Evan is verified.",
"Chen is active.",
"Farah is not trusted.",
"Farah is ready.",
"Gabe is not enabled.",
"Gabe is registered."
] | [
"Whenever person 1 is verified and person 1 is enabled, person 1 is not ready; this applies to all people mentioned.",
"Whenever person 1 is verified and person 1 is not enabled, person 1 is ready; this applies to all people mentioned.",
"Whenever person 1 is registered, person 1 is verified; this applies to al... | Gabe is not ready. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Evan is verified.
Chen is active.
Farah is not trusted.
Farah is ready.
Gabe is not enab... |
v7-r1-calibration-old-000057-calibration-false | v7-r1-calibration-old-000057 | calibration | {
"entities": [
"Evan",
"Farah",
"Chen"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "blue",
"arguments": [
"Farah"
],
"negative": true
},
{
"predicate": "calm... | {
"predicate": "bright",
"arguments": [
"Evan"
],
"negative": true
} | FALSE | 1 | 1 | chain | conditional | 2489c2d3351e4a6a2a263f6ad312ec5eeeb95d32927679081db48d8c93b96aad | d6cdea0617b6466794837beafb7653bbf07acb9494a5c628e1b8f13f481fdb99 | [
"Chen is not calm.",
"Farah is not blue.",
"Chen is kind.",
"Chen is not round.",
"Farah is round.",
"Farah is not kind.",
"Evan is quiet.",
"Chen is blue.",
"Chen is not large.",
"Evan is warm."
] | [
"Whenever person 1 is quiet and person 1 is warm, person 1 is bright; this applies to all people mentioned.",
"Whenever person 1 is quiet and person 1 is not warm, person 1 is not bright; this applies to all people mentioned."
] | Evan is not bright. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is not calm.
Farah is not blue.
Chen is kind.
Chen is not round.
Farah is round.
Fa... |
v7-r1-calibration-multi-4-0004-0-multi-renamed-2 | v7-r1-calibration-multi-4-0004-0 | calibration | {
"entities": [
"Farah",
"Gabe",
"Hana",
"Evan"
],
"facts": [
{
"predicate": "enabled",
"arguments": [
"Gabe"
],
"negative": false
},
{
"predicate": "enabled",
"arguments": [
"Farah"
],
"negative": false
},
{
... | {
"predicate": "ready",
"arguments": [
"Hana"
],
"negative": false
} | TRUE | 2 | 2 | conjunction | canonical | 0c2ad4168f599a20189c3c01553efafa2b96fa3e2430748f80ee437500b44836 | 706cae65cedc1b3d33cb62d3f2a295e30222a06f05857cad6e0a0711fd327dd7 | [
"Evan is not verified.",
"Farah is approved.",
"Gabe is verified.",
"Evan is registered.",
"Farah is registered.",
"Hana is enabled.",
"Farah is enabled.",
"Hana is verified.",
"Gabe is enabled.",
"Gabe is registered.",
"Hana is registered.",
"Evan is enabled."
] | [
"For every person mentioned: if person 1 is registered and person 1 is not verified, then person 1 is not active.",
"For every person mentioned: if person 1 is active, then person 1 is ready.",
"For every person mentioned: if person 1 is registered and person 1 is verified, then person 1 is active.",
"For eve... | Hana is ready. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Evan is not verified.
Farah is approved.
Gabe is verified.
Evan is registered.
Farah is ... |
v7-r1-calibration-old-000121-calibration-unknown | v7-r1-calibration-old-000121 | calibration | {
"entities": [
"Chen",
"Evan",
"Bella",
"Gabe"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Bella"
],
"negative": false
},
{
"predicate": "approved",
"arguments": [
"Bella"
],
"negative": false
},
{
... | {
"predicate": "verified",
"arguments": [
"Chen"
],
"negative": true
} | UNKNOWN | null | 0 | chain | canonical | bfd1bf5ba8c12552156fbd188ce7d471328c5e63c9f4366fd6c8e13b4daccdf9 | 452fc03fe2b109b042d10502e5770d53dc7985bb632c4f2fa96a4e6b8a8a36dc | [
"Chen is not authorized.",
"Chen is not registered.",
"Bella is active.",
"Bella is trusted.",
"Bella is approved.",
"Gabe is enabled.",
"Evan is not enabled.",
"Bella is not ready.",
"Evan is not verified."
] | [
"For every person mentioned: if person 1 is trusted and person 1 is not authorized, then person 1 is not verified.",
"For every person mentioned: if person 1 is trusted and person 1 is authorized, then person 1 is verified.",
"For every person mentioned: if person 1 is registered, then person 1 is active.",
"... | Chen is not verified. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is not authorized.
Chen is not registered.
Bella is active.
Bella is trusted.
Bella... |
v7-r1-calibration-multi-3-0002-2-multi-base-1 | v7-r1-calibration-multi-3-0002-2 | calibration | {
"entities": [
"Chen",
"Dalia",
"Evan"
],
"facts": [
{
"predicate": "kind",
"arguments": [
"Evan"
],
"negative": false
},
{
"predicate": "kind",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": "war... | {
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": false
} | TRUE | 2 | 2 | conjunction | canonical | 4a600256b853bde0d8114e63a83a65f3ecf22589cf3d43225b429d1203f40faf | 706cae65cedc1b3d33cb62d3f2a295e30222a06f05857cad6e0a0711fd327dd7 | [
"Chen is not kind.",
"Dalia is warm.",
"Dalia is kind.",
"Evan is warm.",
"Chen is round.",
"Dalia is round.",
"Evan is round.",
"Chen is large.",
"Evan is kind."
] | [
"For every person mentioned: if person 1 is not bright, then person 1 is not blue.",
"For every person mentioned: if person 1 is round and person 1 is not kind, then person 1 is not bright.",
"For every person mentioned: if person 1 is bright, then person 1 is blue.",
"For every person mentioned: if person 1 ... | Dalia is blue. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is not kind.
Dalia is warm.
Dalia is kind.
Evan is warm.
Chen is round.
Dalia is ro... |
v7-r1-calibration-multi-5-0006-0-multi-base-0 | v7-r1-calibration-multi-5-0006-0 | calibration | {
"entities": [
"Gabe",
"Hana",
"EntityA",
"Bella",
"Chen"
],
"facts": [
{
"predicate": "calm",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "large",
"arguments": [
"EntityA"
],
"negative": false
... | {
"predicate": "kind",
"arguments": [
"Gabe"
],
"negative": false
} | TRUE | 1 | 1 | conjunction | canonical | a7c3a7063a10c710109206fbee24eee6da33d8b70ee347711be84f691e33dfdf | cadd90deb012a4b9628b7ec27362942c3e58b1b395b20e891e38650ad17af718 | [
"Bella is calm.",
"Bella is bright.",
"Gabe is calm.",
"Hana is large.",
"Bella is large.",
"Gabe is quiet.",
"Gabe is large.",
"Hana is quiet.",
"EntityA is calm.",
"Chen is calm.",
"Hana is calm.",
"EntityA is large.",
"Chen is bright.",
"EntityA is bright.",
"Chen is large."
] | [
"For every person mentioned: if person 1 is calm and person 1 is not large, then person 1 is not kind.",
"For every person mentioned: if person 1 is calm and person 1 is large, then person 1 is kind."
] | Gabe is kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is calm.
Bella is bright.
Gabe is calm.
Hana is large.
Bella is large.
Gabe is qui... |
v7-r1-calibration-multi-5-0006-0-multi-support_edit-4 | v7-r1-calibration-multi-5-0006-0 | calibration | {
"entities": [
"Gabe",
"Hana",
"EntityA",
"Bella",
"Chen"
],
"facts": [
{
"predicate": "calm",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "large",
"arguments": [
"EntityA"
],
"negative": false
... | {
"predicate": "kind",
"arguments": [
"Chen"
],
"negative": false
} | TRUE | 1 | 1 | conjunction | canonical | 60330ac912f81f598401bd872508fccaf5d3d739dbbf174f165ad242689c9d89 | cadd90deb012a4b9628b7ec27362942c3e58b1b395b20e891e38650ad17af718 | [
"Bella is calm.",
"Hana is bright.",
"Gabe is calm.",
"Bella is large.",
"Bella is bright.",
"Gabe is quiet.",
"Gabe is large.",
"Hana is quiet.",
"EntityA is calm.",
"Chen is calm.",
"Hana is calm.",
"EntityA is large.",
"Chen is bright.",
"EntityA is bright.",
"Chen is large."
] | [
"For every person mentioned: if person 1 is calm and person 1 is not large, then person 1 is not kind.",
"For every person mentioned: if person 1 is calm and person 1 is large, then person 1 is kind."
] | Chen is kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is calm.
Hana is bright.
Gabe is calm.
Bella is large.
Bella is bright.
Gabe is qu... |
v7-r1-calibration-old-000035-calibration-true | v7-r1-calibration-old-000035 | calibration | {
"entities": [
"Evan",
"EntityA",
"Gabe",
"Dalia"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": false
},
{
... | {
"predicate": "warm",
"arguments": [
"Evan"
],
"negative": false
} | TRUE | 1 | 1 | relational | conditional | 8a1bc578bcd9dddf87aa2727eddf0524254648c12fc9bb818a6599eef89ec68f | 5b711a9c568ac2fd187eae882b1a35a74b5549b53da284a9e55be83fad8e4b23 | [
"EntityA is blue.",
"Dalia is blue.",
"Evan is bright.",
"Dalia is warm.",
"Dalia is calm.",
"Evan is quiet."
] | [
"Whenever person 1 is quiet and person 1 is bright, person 1 is warm; this applies to all people mentioned.",
"Whenever person 1 is quiet and person 1 is not bright, person 1 is not warm; this applies to all people mentioned."
] | Evan is warm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
EntityA is blue.
Dalia is blue.
Evan is bright.
Dalia is warm.
Dalia is calm.
Evan is qu... |
v7-r1-calibration-multi-4-0004-1-multi-base-3 | v7-r1-calibration-multi-4-0004-1 | calibration | {
"entities": [
"Evan",
"Farah",
"Gabe",
"Hana"
],
"facts": [
{
"predicate": "authorized",
"arguments": [
"Evan"
],
"negative": false
},
{
"predicate": "verified",
"arguments": [
"Farah"
],
"negative": false
},
{
... | {
"predicate": "enabled",
"arguments": [
"Hana"
],
"negative": false
} | UNKNOWN | null | 1 | conjunction | canonical | c57fdc2676dc721b631c6aa0e0f0a7407e487e12712ad6cd42b33f658e46dbd8 | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Farah is verified.",
"Evan is authorized.",
"Farah is not ready.",
"Gabe is not ready.",
"Gabe is verified.",
"Hana is trusted.",
"Farah is trusted.",
"Evan is ready.",
"Hana is verified.",
"Gabe is trusted.",
"Hana is authorized.",
"Evan is verified."
] | [
"For every person mentioned: if person 1 is verified and person 1 is not ready, then person 1 is not enabled.",
"For every person mentioned: if person 1 is verified and person 1 is ready, then person 1 is enabled."
] | Hana is enabled. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Farah is verified.
Evan is authorized.
Farah is not ready.
Gabe is not ready.
Gabe is ve... |
v7-r1-calibration-multi-4-0002-1-multi-support_edit-0 | v7-r1-calibration-multi-4-0002-1 | calibration | {
"entities": [
"Chen",
"Dalia",
"Evan",
"Farah"
],
"facts": [
{
"predicate": "verified",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "active",
"arguments": [
"Farah"
],
"negative": true
},
{
... | {
"predicate": "registered",
"arguments": [
"Chen"
],
"negative": false
} | UNKNOWN | null | 2 | conjunction | canonical | 7d097fef2d15e0f7e877c80b78d5aed97c9731507258971a64a89493301449e4 | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Evan is enabled.",
"Chen is authorized.",
"Dalia is enabled.",
"Evan is active.",
"Farah is not active.",
"Evan is authorized.",
"Dalia is not active.",
"Dalia is verified.",
"Chen is enabled.",
"Farah is authorized.",
"Chen is verified.",
"Farah is enabled."
] | [
"For every person mentioned: if person 1 is enabled and person 1 is active, then person 1 is ready.",
"For every person mentioned: if person 1 is enabled and person 1 is not active, then person 1 is not ready.",
"For every person mentioned: if person 1 is ready, then person 1 is registered.",
"For every perso... | Chen is registered. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Evan is enabled.
Chen is authorized.
Dalia is enabled.
Evan is active.
Farah is not acti... |
v7-r1-calibration-old-000134-calibration-true | v7-r1-calibration-old-000134 | calibration | {
"entities": [
"Gabe",
"Dalia",
"Bella",
"Chen"
],
"facts": [
{
"predicate": "approved",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "approved",
"arguments": [
"Gabe"
],
"negative": true
},
{
... | {
"predicate": "enabled",
"arguments": [
"Bella"
],
"negative": true
} | TRUE | 3 | 3 | relational | canonical | 664fc3cce5548dacab8fac87f8f40d1f7e4f8c75be392411a5fd7d6b93d47797 | ca1f529f2d76e8e33e8e05b33259f6d92c1740d6605cd6d61b5f2c5d1ef0eb0c | [
"Gabe is not approved.",
"Chen is enabled.",
"Dalia is not enabled.",
"Gabe is not trusted.",
"Bella follows Dalia.",
"Bella is verified.",
"Dalia knows Gabe.",
"Chen is approved."
] | [
"For every person mentioned: if person 1 follows person 2 and person 2 is registered, then person 1 is not ready.",
"For every person mentioned: if person 1 knows person 2 and person 2 is not trusted, then person 1 is registered.",
"For every person mentioned: if person 1 is not ready and person 1 is verified, ... | Bella is not enabled. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is not approved.
Chen is enabled.
Dalia is not enabled.
Gabe is not trusted.
Bella ... |
v7-r1-calibration-multi-4-0003-1-multi-base-0 | v7-r1-calibration-multi-4-0003-1 | calibration | {
"entities": [
"Dalia",
"Evan",
"Farah",
"Gabe"
],
"facts": [
{
"predicate": "ready",
"arguments": [
"Farah"
],
"negative": false
},
{
"predicate": "ready",
"arguments": [
"Evan"
],
"negative": false
},
{
"p... | {
"predicate": "trusted",
"arguments": [
"Dalia"
],
"negative": true
} | FALSE | 2 | 2 | signed_chain | canonical | fb87203e00c26ce577d61a83375d2ca6b720722d78874234b429fddf6837880b | 1ca429712ede2b26a785e857863d1272b75704c0b2462aae6c4c002415bb4292 | [
"Dalia is ready.",
"Gabe is active.",
"Farah is registered.",
"Gabe is not authorized.",
"Evan is not authorized.",
"Dalia is authorized.",
"Farah is active.",
"Dalia is registered.",
"Farah is ready.",
"Evan is active.",
"Gabe is ready.",
"Evan is ready."
] | [
"For every person mentioned: if person 1 is authorized, then person 1 is approved.",
"For every person mentioned: if person 1 is not approved, then person 1 is not trusted.",
"For every person mentioned: if person 1 is approved, then person 1 is trusted.",
"For every person mentioned: if person 1 is not autho... | Dalia is not trusted. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is ready.
Gabe is active.
Farah is registered.
Gabe is not authorized.
Evan is not... |
v7-r1-calibration-old-000006-calibration-true | v7-r1-calibration-old-000006 | calibration | {
"entities": [
"Farah",
"Hana",
"Dalia"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Farah"
],
"negative": true
},
{
"predicate": "approved",
"arguments": [
"Hana"
],
"negative": true
},
{
"predicate":... | {
"predicate": "active",
"arguments": [
"Hana"
],
"negative": false
} | TRUE | 2 | 2 | relational | conditional | 472e7d58df5e2fba6f8e104a951ded8838bbab6a4f6b0525f501b09f2ebdb2d7 | 44683fac69fbc06ccbf74f16e30207e66c2ec7d4a3a61270b430a85a65b794c8 | [
"Hana knows Farah.",
"Hana is not approved.",
"Farah is registered.",
"Farah is not trusted.",
"Dalia is ready.",
"Dalia is not authorized.",
"Farah is not active.",
"Farah is not verified."
] | [
"Whenever person 1 is not verified and person 1 is not approved, person 1 is active; this applies to all people mentioned.",
"Whenever person 1 is not verified and person 1 is approved, person 1 is not active; this applies to all people mentioned.",
"Whenever person 1 knows person 2 and person 2 is registered, ... | Hana is active. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Hana knows Farah.
Hana is not approved.
Farah is registered.
Farah is not trusted.
Dalia... |
v7-r1-calibration-multi-5-0006-2-multi-base-2 | v7-r1-calibration-multi-5-0006-2 | calibration | {
"entities": [
"Gabe",
"Hana",
"EntityA",
"Bella",
"Chen"
],
"facts": [
{
"predicate": "warm",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "quiet",
"arguments": [
"Gabe"
],
"negative": false
},... | {
"predicate": "kind",
"arguments": [
"EntityA"
],
"negative": false
} | UNKNOWN | null | 0 | conjunction | canonical | 81d6be4dd3b902f226f71a395a79afa4c927a53b98c91160ae105343bfb4310e | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Hana is warm.",
"EntityA is blue.",
"Gabe is blue.",
"Chen is blue.",
"Hana is blue.",
"Chen is quiet.",
"Hana is quiet.",
"Bella is blue.",
"Bella is warm.",
"Bella is quiet.",
"EntityA is quiet.",
"EntityA is warm.",
"Gabe is quiet.",
"Gabe is warm.",
"Chen is warm."
] | [
"For every person mentioned: if person 1 is quiet and person 1 is not bright, then person 1 is not kind.",
"For every person mentioned: if person 1 is quiet and person 1 is bright, then person 1 is kind."
] | EntityA is kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Hana is warm.
EntityA is blue.
Gabe is blue.
Chen is blue.
Hana is blue.
Chen is quiet.
... |
v7-r1-calibration-old-000039-calibration-true | v7-r1-calibration-old-000039 | calibration | {
"entities": [
"Evan",
"Hana",
"Gabe",
"Farah"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Farah"
],
"negative": false
},
{
"predicate": "approved",
"arguments": [
"Evan"
],
"negative": false
},
{
... | {
"predicate": "registered",
"arguments": [
"Hana"
],
"negative": true
} | TRUE | 2 | 2 | relational | conditional | b210b6eafdb7a524b51a0d4aef0de01dd208af6b2c9eb9a9a3fe9902301cfa77 | 44683fac69fbc06ccbf74f16e30207e66c2ec7d4a3a61270b430a85a65b794c8 | [
"Farah is active.",
"Evan is approved.",
"Hana knows Evan.",
"Farah is not approved.",
"Gabe is not approved.",
"Gabe is not registered.",
"Evan is not trusted.",
"Evan is not enabled.",
"Hana is not authorized.",
"Gabe is trusted."
] | [
"Whenever person 1 knows person 2 and person 2 is not enabled, person 1 is not verified; this applies to all people mentioned.",
"Whenever person 1 is not verified and person 1 is authorized, person 1 is registered; this applies to all people mentioned.",
"Whenever person 1 is not verified and person 1 is not a... | Hana is not registered. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Farah is active.
Evan is approved.
Hana knows Evan.
Farah is not approved.
Gabe is not a... |
v7-r1-calibration-old-000108-calibration-true | v7-r1-calibration-old-000108 | calibration | {
"entities": [
"Chen",
"Gabe",
"EntityA"
],
"facts": [
{
"predicate": "authorized",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "knows",
"arguments": [
"Gabe",
"Chen"
],
"negative": false
},
... | {
"predicate": "active",
"arguments": [
"Gabe"
],
"negative": false
} | TRUE | 2 | 2 | relational | canonical | eb1cf2ebcff0eb18873f9f82e39171ee0eb6aeed37059f21d3634420f25e2743 | 44683fac69fbc06ccbf74f16e30207e66c2ec7d4a3a61270b430a85a65b794c8 | [
"Gabe is not verified.",
"Chen is not ready.",
"Gabe knows Chen.",
"Chen is not registered.",
"Chen is trusted.",
"EntityA is authorized.",
"EntityA is not trusted."
] | [
"For every person mentioned: if person 1 is enabled and person 1 is verified, then person 1 is not active.",
"For every person mentioned: if person 1 is enabled and person 1 is not verified, then person 1 is active.",
"For every person mentioned: if person 1 knows person 2 and person 2 is not ready, then person... | Gabe is active. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is not verified.
Chen is not ready.
Gabe knows Chen.
Chen is not registered.
Chen i... |
v7-r1-calibration-multi-3-0002-0-multi-support_edit-0 | v7-r1-calibration-multi-3-0002-0 | calibration | {
"entities": [
"Chen",
"Dalia",
"Evan"
],
"facts": [
{
"predicate": "quiet",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": true
},
{
"predicate": "cal... | {
"predicate": "warm",
"arguments": [
"Chen"
],
"negative": false
} | UNKNOWN | null | 1 | conjunction | canonical | 2578b57e26a6edf81c658f41aff1777e48664911447144f540a0637220888a67 | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Dalia is calm.",
"Evan is kind.",
"Evan is blue.",
"Evan is calm.",
"Chen is kind.",
"Dalia is not blue.",
"Chen is quiet.",
"Dalia is quiet.",
"Chen is calm."
] | [
"For every person mentioned: if person 1 is calm and person 1 is blue, then person 1 is warm.",
"For every person mentioned: if person 1 is calm and person 1 is not blue, then person 1 is not warm."
] | Chen is warm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is calm.
Evan is kind.
Evan is blue.
Evan is calm.
Chen is kind.
Dalia is not blue... |
v7-r1-calibration-old-000102-calibration-false | v7-r1-calibration-old-000102 | calibration | {
"entities": [
"Chen",
"Evan",
"Dalia"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": "approved",
"arguments": [
"Chen"
],
"negative": true
},
{
"predicate":... | {
"predicate": "verified",
"arguments": [
"Chen"
],
"negative": true
} | FALSE | 1 | 1 | relational | conditional | 109cfa95f471bbe30062fdd7775c001da9ed6243eb7d75ae3ca5e8aebec795dd | 5b711a9c568ac2fd187eae882b1a35a74b5549b53da284a9e55be83fad8e4b23 | [
"Chen is not ready.",
"Dalia is trusted.",
"Dalia is active.",
"Dalia is not verified.",
"Chen is not approved.",
"Dalia is authorized."
] | [
"Whenever person 1 is not approved and person 1 is not ready, person 1 is verified; this applies to all people mentioned.",
"Whenever person 1 is not approved and person 1 is ready, person 1 is not verified; this applies to all people mentioned."
] | Chen is not verified. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is not ready.
Dalia is trusted.
Dalia is active.
Dalia is not verified.
Chen is not... |
v7-r1-calibration-multi-3-0001-0-multi-reordered-0 | v7-r1-calibration-multi-3-0001-0 | calibration | {
"entities": [
"Bella",
"Chen",
"Dalia"
],
"facts": [
{
"predicate": "bright",
"arguments": [
"Chen"
],
"negative": false
},
{
"predicate": "large",
"arguments": [
"Bella"
],
"negative": false
},
{
"predicate": ... | {
"predicate": "blue",
"arguments": [
"Bella"
],
"negative": true
} | UNKNOWN | null | 1 | conjunction | canonical | 3999e8ad2bf0aabef2b4e7fff7a6339cd6f92e6f2f202e7ed2bb550c2230785a | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Chen is round.",
"Chen is calm.",
"Chen is bright.",
"Dalia is large.",
"Bella is large.",
"Bella is round.",
"Bella is bright.",
"Dalia is round.",
"Dalia is not calm."
] | [
"For every person mentioned: if person 1 is round and person 1 is not calm, then person 1 is not blue.",
"For every person mentioned: if person 1 is round and person 1 is calm, then person 1 is blue."
] | Bella is not blue. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is round.
Chen is calm.
Chen is bright.
Dalia is large.
Bella is large.
Bella is ro... |
v7-r1-calibration-multi-5-0005-1-multi-base-3 | v7-r1-calibration-multi-5-0005-1 | calibration | {
"entities": [
"Farah",
"Gabe",
"Hana",
"EntityA",
"Bella"
],
"facts": [
{
"predicate": "registered",
"arguments": [
"Farah"
],
"negative": false
},
{
"predicate": "registered",
"arguments": [
"Hana"
],
"negative": fa... | {
"predicate": "enabled",
"arguments": [
"EntityA"
],
"negative": true
} | FALSE | 2 | 2 | conjunction | canonical | 0d1a2d5a8e5a68eea46a6d203c463f697382c3574c7585ef3b18b761a493e861 | 706cae65cedc1b3d33cb62d3f2a295e30222a06f05857cad6e0a0711fd327dd7 | [
"Farah is ready.",
"Hana is authorized.",
"Hana is registered.",
"Bella is authorized.",
"Gabe is authorized.",
"Hana is verified.",
"Bella is ready.",
"EntityA is authorized.",
"EntityA is ready.",
"Farah is authorized.",
"Gabe is verified.",
"Farah is registered.",
"EntityA is verified.",
... | [
"For every person mentioned: if person 1 is trusted, then person 1 is enabled.",
"For every person mentioned: if person 1 is authorized and person 1 is not verified, then person 1 is not trusted.",
"For every person mentioned: if person 1 is not trusted, then person 1 is not enabled.",
"For every person menti... | EntityA is not enabled. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Farah is ready.
Hana is authorized.
Hana is registered.
Bella is authorized.
Gabe is aut... |
v7-r1-calibration-multi-3-0001-1-multi-reordered-2 | v7-r1-calibration-multi-3-0001-1 | calibration | {
"entities": [
"Bella",
"Chen",
"Dalia"
],
"facts": [
{
"predicate": "large",
"arguments": [
"Bella"
],
"negative": false
},
{
"predicate": "warm",
"arguments": [
"Bella"
],
"negative": true
},
{
"predicate": "l... | {
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": true
} | FALSE | 2 | 2 | conjunction | canonical | 3784f3bce3adccf8b33762bf287eb20bd96d5b0d9fb3fed260e6c127eeff4240 | 706cae65cedc1b3d33cb62d3f2a295e30222a06f05857cad6e0a0711fd327dd7 | [
"Bella is kind.",
"Chen is quiet.",
"Bella is large.",
"Chen is large.",
"Chen is kind.",
"Bella is not warm.",
"Dalia is kind.",
"Dalia is warm.",
"Dalia is large."
] | [
"For every person mentioned: if person 1 is not round, then person 1 is not blue.",
"For every person mentioned: if person 1 is kind and person 1 is not warm, then person 1 is not round.",
"For every person mentioned: if person 1 is round, then person 1 is blue.",
"For every person mentioned: if person 1 is k... | Dalia is not blue. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is kind.
Chen is quiet.
Bella is large.
Chen is large.
Chen is kind.
Bella is not ... |
v7-r1-calibration-multi-5-0003-2-multi-base-2 | v7-r1-calibration-multi-5-0003-2 | calibration | {
"entities": [
"Dalia",
"Evan",
"Farah",
"Gabe",
"Hana"
],
"facts": [
{
"predicate": "bright",
"arguments": [
"Farah"
],
"negative": false
},
{
"predicate": "quiet",
"arguments": [
"Farah"
],
"negative": false
},
... | {
"predicate": "warm",
"arguments": [
"Farah"
],
"negative": true
} | UNKNOWN | null | 2 | signed_chain | canonical | c0417e02ea0074f94ad5b780e9bfffc9f82246764521b10f70e64c3cb1519715 | cdfd8d92931b901caba0f41203af2efb0f3e650038aa30522d44024f1a911cb1 | [
"Gabe is large.",
"Dalia is large.",
"Hana is calm.",
"Farah is large.",
"Gabe is quiet.",
"Farah is quiet.",
"Evan is calm.",
"Gabe is bright.",
"Hana is bright.",
"Evan is large.",
"Dalia is calm.",
"Evan is quiet.",
"Dalia is quiet.",
"Farah is bright.",
"Hana is large."
] | [
"For every person mentioned: if person 1 is calm, then person 1 is kind.",
"For every person mentioned: if person 1 is kind, then person 1 is warm.",
"For every person mentioned: if person 1 is not calm, then person 1 is not kind.",
"For every person mentioned: if person 1 is not kind, then person 1 is not wa... | Farah is not warm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is large.
Dalia is large.
Hana is calm.
Farah is large.
Gabe is quiet.
Farah is qui... |
v7-r1-calibration-old-000131-calibration-true | v7-r1-calibration-old-000131 | calibration | {
"entities": [
"EntityA",
"Dalia",
"Chen"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"EntityA"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Chen"
],
"negative": true
},
{
"predicate":... | {
"predicate": "large",
"arguments": [
"Dalia"
],
"negative": false
} | TRUE | 2 | 2 | relational | conditional | 7a1dc2d1706b5d92ccad5058ea73ec5ace005f2b2d78c38dbebc52cf89293560 | 44683fac69fbc06ccbf74f16e30207e66c2ec7d4a3a61270b430a85a65b794c8 | [
"Dalia knows EntityA.",
"Chen is not kind.",
"Chen is not bright.",
"Chen is quiet.",
"EntityA is quiet.",
"EntityA is not blue.",
"Dalia is not round."
] | [
"Whenever person 1 is not quiet and person 1 is round, person 1 is not large; this applies to all people mentioned.",
"Whenever person 1 knows person 2 and person 2 is not blue, person 1 is not quiet; this applies to all people mentioned.",
"Whenever person 1 is not quiet and person 1 is not round, person 1 is ... | Dalia is large. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia knows EntityA.
Chen is not kind.
Chen is not bright.
Chen is quiet.
EntityA is qui... |
v7-r1-calibration-old-000061-calibration-false | v7-r1-calibration-old-000061 | calibration | {
"entities": [
"Bella",
"Dalia",
"EntityA"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Bella"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate"... | {
"predicate": "kind",
"arguments": [
"Bella"
],
"negative": true
} | FALSE | 3 | 3 | chain | conditional | fec2ad614785aa1c31a3913a3e81c575694dc8cb668b34b9897dba3da1fc1a3d | 8628c5aad256c8edba6f7f2657c0483db19285b433a26e9bbaccd959071fc344 | [
"Bella is round.",
"Bella is not blue.",
"EntityA is not calm.",
"Dalia is not kind.",
"Dalia is not quiet.",
"Dalia is bright.",
"EntityA is warm."
] | [
"Whenever person 1 is not bright, person 1 is calm; this applies to all people mentioned.",
"Whenever person 1 is calm and person 1 is not round, person 1 is not kind; this applies to all people mentioned.",
"Whenever person 1 is calm and person 1 is round, person 1 is kind; this applies to all people mentioned... | Bella is not kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is round.
Bella is not blue.
EntityA is not calm.
Dalia is not kind.
Dalia is not ... |
v7-r1-calibration-old-000061-calibration-unknown | v7-r1-calibration-old-000061 | calibration | {
"entities": [
"Bella",
"Dalia",
"EntityA"
],
"facts": [
{
"predicate": "bright",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": "calm",
"arguments": [
"EntityA"
],
"negative": true
},
{
"predicat... | {
"predicate": "kind",
"arguments": [
"Bella"
],
"negative": true
} | UNKNOWN | null | 0 | chain | conditional | 57da86c7bdbc8ca813d8ac4ef9d8425f280650a0c36695014ff9bccca87ee872 | 452fc03fe2b109b042d10502e5770d53dc7985bb632c4f2fa96a4e6b8a8a36dc | [
"Bella is blue.",
"Dalia is bright.",
"Dalia is not kind.",
"Dalia is not quiet.",
"Bella is not round.",
"EntityA is not calm.",
"EntityA is warm."
] | [
"Whenever person 1 is not bright, person 1 is calm; this applies to all people mentioned.",
"Whenever person 1 is calm and person 1 is not round, person 1 is not kind; this applies to all people mentioned.",
"Whenever person 1 is calm and person 1 is round, person 1 is kind; this applies to all people mentioned... | Bella is not kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is blue.
Dalia is bright.
Dalia is not kind.
Dalia is not quiet.
Bella is not roun... |
v7-r1-calibration-old-000114-calibration-false | v7-r1-calibration-old-000114 | calibration | {
"entities": [
"Farah",
"Dalia",
"Hana",
"Bella"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Bella"
],
"negative": false
},
{
"predicate": "active",
"arguments": [
"Dalia"
],
"negative": false
},
{
... | {
"predicate": "enabled",
"arguments": [
"Farah"
],
"negative": true
} | FALSE | 1 | 1 | chain | canonical | a02de9dda2af20755a44d4f57fcc99f2dcc3cb06a221ac3abcf1d58def776e0b | d6cdea0617b6466794837beafb7653bbf07acb9494a5c628e1b8f13f481fdb99 | [
"Dalia is approved.",
"Farah is registered.",
"Farah is not verified.",
"Bella is ready.",
"Dalia is active.",
"Bella is active."
] | [
"For every person mentioned: if person 1 is registered and person 1 is not verified, then person 1 is enabled.",
"For every person mentioned: if person 1 is registered and person 1 is verified, then person 1 is not enabled."
] | Farah is not enabled. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is approved.
Farah is registered.
Farah is not verified.
Bella is ready.
Dalia is ... |
v7-r1-calibration-multi-3-0006-0-multi-base-1 | v7-r1-calibration-multi-3-0006-0 | calibration | {
"entities": [
"Gabe",
"Hana",
"EntityA"
],
"facts": [
{
"predicate": "authorized",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "authorized",
"arguments": [
"Hana"
],
"negative": false
},
{
"... | {
"predicate": "verified",
"arguments": [
"Hana"
],
"negative": false
} | TRUE | 2 | 2 | conjunction | canonical | 2a10d459f2af7e238f0ef2ce79f9835ca54710200af978316c2b643ccaffbf14 | 706cae65cedc1b3d33cb62d3f2a295e30222a06f05857cad6e0a0711fd327dd7 | [
"Hana is ready.",
"EntityA is ready.",
"EntityA is active.",
"Hana is trusted.",
"Gabe is ready.",
"Gabe is authorized.",
"Gabe is trusted.",
"EntityA is authorized.",
"Hana is authorized."
] | [
"For every person mentioned: if person 1 is not registered, then person 1 is not verified.",
"For every person mentioned: if person 1 is registered, then person 1 is verified.",
"For every person mentioned: if person 1 is authorized and person 1 is ready, then person 1 is registered.",
"For every person menti... | Hana is verified. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Hana is ready.
EntityA is ready.
EntityA is active.
Hana is trusted.
Gabe is ready.
Gabe... |
v7-r1-calibration-multi-5-0005-0-multi-reordered-0 | v7-r1-calibration-multi-5-0005-0 | calibration | {
"entities": [
"Farah",
"Gabe",
"Hana",
"EntityA",
"Bella"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "quiet",
"arguments": [
"EntityA"
],
"negative": false
... | {
"predicate": "round",
"arguments": [
"Farah"
],
"negative": true
} | FALSE | 1 | 1 | conjunction | canonical | 81444f0dfa2ae819cf9268469eacbdc438ed48ae1cf90cb9a91c7dc8dc35a62e | cadd90deb012a4b9628b7ec27362942c3e58b1b395b20e891e38650ad17af718 | [
"EntityA is blue.",
"EntityA is not bright.",
"Gabe is warm.",
"EntityA is quiet.",
"Bella is blue.",
"Hana is blue.",
"Hana is not bright.",
"Farah is blue.",
"Bella is bright.",
"Bella is quiet.",
"Gabe is not bright.",
"Hana is quiet.",
"Farah is bright.",
"Farah is quiet.",
"Gabe is ... | [
"For every person mentioned: if person 1 is blue and person 1 is bright, then person 1 is round.",
"For every person mentioned: if person 1 is blue and person 1 is not bright, then person 1 is not round."
] | Farah is not round. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
EntityA is blue.
EntityA is not bright.
Gabe is warm.
EntityA is quiet.
Bella is blue.
H... |
v7-r1-calibration-old-000001-calibration-unknown | v7-r1-calibration-old-000001 | calibration | {
"entities": [
"Chen",
"Gabe",
"Hana",
"Dalia"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Gabe"
],
"negative": false
},
{
"predicate": "authorized",
"arguments": [
"Gabe"
],
"negative": true
},
{
... | {
"predicate": "ready",
"arguments": [
"Chen"
],
"negative": false
} | UNKNOWN | null | 0 | chain | conditional | cae175089fb8bfb4440cc40366262e4e25277eff8996a3383b8202606388a12e | 452fc03fe2b109b042d10502e5770d53dc7985bb632c4f2fa96a4e6b8a8a36dc | [
"Dalia is ready.",
"Hana is authorized.",
"Gabe is not authorized.",
"Chen is enabled.",
"Gabe is active.",
"Chen is approved."
] | [
"Whenever person 1 is not verified and person 1 is not enabled, person 1 is not ready; this applies to all people mentioned.",
"Whenever person 1 is not approved, person 1 is not verified; this applies to all people mentioned.",
"Whenever person 1 is not verified and person 1 is enabled, person 1 is ready; this... | Chen is ready. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is ready.
Hana is authorized.
Gabe is not authorized.
Chen is enabled.
Gabe is act... |
v7-r1-calibration-multi-5-0004-0-multi-base-4 | v7-r1-calibration-multi-5-0004-0 | calibration | {
"entities": [
"Evan",
"Farah",
"Gabe",
"Hana",
"EntityA"
],
"facts": [
{
"predicate": "trusted",
"arguments": [
"Hana"
],
"negative": false
},
{
"predicate": "enabled",
"arguments": [
"Farah"
],
"negative": false
... | {
"predicate": "authorized",
"arguments": [
"EntityA"
],
"negative": false
} | UNKNOWN | null | 1 | conjunction | canonical | aa7f180090ec9a30ac2c1f36b60ba558d64e1b6b0f5e7137bc16d2cf8d5acb2d | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Gabe is enabled.",
"Farah is approved.",
"Gabe is not trusted.",
"Evan is approved.",
"Hana is approved.",
"Gabe is approved.",
"Hana is enabled.",
"Evan is trusted.",
"EntityA is active.",
"EntityA is approved.",
"EntityA is enabled.",
"Evan is active.",
"Farah is trusted.",
"Farah is en... | [
"For every person mentioned: if person 1 is approved and person 1 is trusted, then person 1 is authorized.",
"For every person mentioned: if person 1 is approved and person 1 is not trusted, then person 1 is not authorized."
] | EntityA is authorized. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is enabled.
Farah is approved.
Gabe is not trusted.
Evan is approved.
Hana is appro... |
v7-r1-calibration-multi-4-0002-2-multi-base-0 | v7-r1-calibration-multi-4-0002-2 | calibration | {
"entities": [
"Chen",
"Dalia",
"Evan",
"Farah"
],
"facts": [
{
"predicate": "enabled",
"arguments": [
"Evan"
],
"negative": false
},
{
"predicate": "verified",
"arguments": [
"Farah"
],
"negative": false
},
{
... | {
"predicate": "approved",
"arguments": [
"Chen"
],
"negative": false
} | TRUE | 1 | 1 | conjunction | canonical | 54a9b0b1584433a09337d265e72b7166369efb0f3dcdf2ec62b4221ba1f413e6 | cadd90deb012a4b9628b7ec27362942c3e58b1b395b20e891e38650ad17af718 | [
"Farah is verified.",
"Evan is verified.",
"Dalia is enabled.",
"Farah is enabled.",
"Chen is trusted.",
"Dalia is verified.",
"Dalia is active.",
"Chen is verified.",
"Evan is active.",
"Evan is enabled.",
"Farah is active.",
"Chen is active."
] | [
"For every person mentioned: if person 1 is verified and person 1 is not trusted, then person 1 is not approved.",
"For every person mentioned: if person 1 is verified and person 1 is trusted, then person 1 is approved."
] | Chen is approved. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Farah is verified.
Evan is verified.
Dalia is enabled.
Farah is enabled.
Chen is trusted... |
v7-r1-calibration-old-000066-calibration-true | v7-r1-calibration-old-000066 | calibration | {
"entities": [
"Gabe",
"Dalia",
"Hana",
"Evan"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Gabe"
],
"negative": false
},
{
"predicate": "bright",
"arguments": [
"Hana"
],
"negative": true
},
{
"pred... | {
"predicate": "kind",
"arguments": [
"Hana"
],
"negative": true
} | TRUE | 3 | 3 | relational | conditional | 7cd8b26d182be277482dc4e53f4ea6031aa8f1759fdb517c47cae940da2058d3 | ca1f529f2d76e8e33e8e05b33259f6d92c1740d6605cd6d61b5f2c5d1ef0eb0c | [
"Hana supports Dalia.",
"Dalia is not warm.",
"Dalia supports Gabe.",
"Gabe is blue.",
"Gabe is not kind.",
"Gabe is not large.",
"Evan is kind.",
"Hana is not bright."
] | [
"Whenever person 1 is not calm and person 1 is not bright, person 1 is not kind; this applies to all people mentioned.",
"Whenever person 1 supports person 2 and person 2 is not round, person 1 is not calm; this applies to all people mentioned.",
"Whenever person 1 is not calm and person 1 is bright, person 1 i... | Hana is not kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Hana supports Dalia.
Dalia is not warm.
Dalia supports Gabe.
Gabe is blue.
Gabe is not k... |
v7-r1-calibration-old-000143-calibration-unknown | v7-r1-calibration-old-000143 | calibration | {
"entities": [
"EntityA",
"Evan",
"Gabe",
"Hana"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Evan"
],
"negative": false
},
{
"predicate": "calm",
"arguments": [
"Gabe"
],
"negative": true
},
{
"pred... | {
"predicate": "round",
"arguments": [
"Evan"
],
"negative": true
} | UNKNOWN | null | 0 | relational | conditional | 5fcceb9e3aaef1d7033641e59bb5192914292e008bb893f2c0938ce5930cb655 | 999d85ff01ce586cf52d04fd5e2d4008e5d57dcf3fdbd38c3b52abceb1cc69b3 | [
"EntityA is not bright.",
"Hana is not warm.",
"Gabe is large.",
"Evan is blue.",
"EntityA is not kind.",
"Gabe is not kind.",
"Gabe is not calm.",
"Hana is not quiet.",
"EntityA is quiet.",
"Evan knows EntityA."
] | [
"Whenever person 1 is kind and person 1 is not blue, person 1 is round; this applies to all people mentioned.",
"Whenever person 1 is kind and person 1 is blue, person 1 is not round; this applies to all people mentioned.",
"Whenever person 1 knows person 2 and person 2 is bright, person 1 is kind; this applies... | Evan is not round. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
EntityA is not bright.
Hana is not warm.
Gabe is large.
Evan is blue.
EntityA is not kin... |
v7-r1-calibration-old-000154-calibration-unknown | v7-r1-calibration-old-000154 | calibration | {
"entities": [
"Evan",
"Dalia",
"EntityA",
"Bella"
],
"facts": [
{
"predicate": "authorized",
"arguments": [
"EntityA"
],
"negative": true
},
{
"predicate": "authorized",
"arguments": [
"Evan"
],
"negative": false
},
... | {
"predicate": "ready",
"arguments": [
"Evan"
],
"negative": false
} | UNKNOWN | null | 2 | relational | canonical | 1b7fa744c7b3ea598be89ed65136a646623e398cffee5a36e330994a36a0f8bb | 999d85ff01ce586cf52d04fd5e2d4008e5d57dcf3fdbd38c3b52abceb1cc69b3 | [
"Bella is not ready.",
"EntityA is not verified.",
"Dalia supports Evan.",
"Dalia is not registered.",
"Evan is authorized.",
"EntityA is ready.",
"EntityA is not authorized.",
"Evan is not trusted.",
"Evan is verified.",
"EntityA is enabled."
] | [
"For every person mentioned: if person 1 is active and person 1 is registered, then person 1 is not ready.",
"For every person mentioned: if person 1 is active and person 1 is not registered, then person 1 is ready.",
"For every person mentioned: if person 1 supports person 2 and person 2 is not trusted, then p... | Evan is ready. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is not ready.
EntityA is not verified.
Dalia supports Evan.
Dalia is not registere... |
v7-r1-calibration-old-000012-calibration-false | v7-r1-calibration-old-000012 | calibration | {
"entities": [
"Evan",
"Hana",
"Farah",
"Chen"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Farah"
],
"negative": false
},
{
"predicate": "bright",
"arguments": [
"Chen"
],
"negative": true
},
{
"pre... | {
"predicate": "large",
"arguments": [
"Evan"
],
"negative": true
} | FALSE | 2 | 2 | chain | conditional | 7ca45d15c173d13b753ea2a4368bc8ae6691ebafcd6aab31c157ba3f919d8978 | 90ced93ad4f5e7b3ebf82a3a74e62256c648008278ba895982d502605377148d | [
"Chen is not bright.",
"Chen is warm.",
"Hana is kind.",
"Farah is kind.",
"Farah is blue.",
"Hana is large.",
"Hana is not warm.",
"Evan is calm.",
"Chen is not large.",
"Evan is quiet."
] | [
"Whenever person 1 is round and person 1 is calm, person 1 is large; this applies to all people mentioned.",
"Whenever person 1 is round and person 1 is not calm, person 1 is not large; this applies to all people mentioned.",
"Whenever person 1 is quiet, person 1 is round; this applies to all people mentioned."... | Evan is not large. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is not bright.
Chen is warm.
Hana is kind.
Farah is kind.
Farah is blue.
Hana is la... |
v7-r1-calibration-old-000156-calibration-unknown | v7-r1-calibration-old-000156 | calibration | {
"entities": [
"Bella",
"Dalia",
"Hana"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Hana"
],
"negative": false
},
{
"predicate": "bright",
"arguments": [
"Bella"
],
"negative": true
},
{
"predicate": "c... | {
"predicate": "warm",
"arguments": [
"Dalia"
],
"negative": false
} | UNKNOWN | null | 2 | chain | conditional | 9f8080b562bb348d07ccb33dd664d5c0070a0194e50b87fb83b06e745a0e79a7 | 452fc03fe2b109b042d10502e5770d53dc7985bb632c4f2fa96a4e6b8a8a36dc | [
"Bella is not bright.",
"Hana is blue.",
"Bella is quiet.",
"Dalia is not calm.",
"Dalia is large.",
"Hana is calm.",
"Hana is not kind."
] | [
"Whenever person 1 is not round and person 1 is not quiet, person 1 is warm; this applies to all people mentioned.",
"Whenever person 1 is not round and person 1 is quiet, person 1 is not warm; this applies to all people mentioned.",
"Whenever person 1 is not bright, person 1 is not round; this applies to all p... | Dalia is warm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Bella is not bright.
Hana is blue.
Bella is quiet.
Dalia is not calm.
Dalia is large.
Ha... |
v7-r1-calibration-multi-5-0003-2-multi-base-1 | v7-r1-calibration-multi-5-0003-2 | calibration | {
"entities": [
"Dalia",
"Evan",
"Farah",
"Gabe",
"Hana"
],
"facts": [
{
"predicate": "bright",
"arguments": [
"Farah"
],
"negative": false
},
{
"predicate": "quiet",
"arguments": [
"Farah"
],
"negative": false
},
... | {
"predicate": "warm",
"arguments": [
"Evan"
],
"negative": true
} | FALSE | 2 | 2 | signed_chain | canonical | c0417e02ea0074f94ad5b780e9bfffc9f82246764521b10f70e64c3cb1519715 | 1ca429712ede2b26a785e857863d1272b75704c0b2462aae6c4c002415bb4292 | [
"Gabe is large.",
"Dalia is large.",
"Hana is calm.",
"Farah is large.",
"Gabe is quiet.",
"Farah is quiet.",
"Evan is calm.",
"Gabe is bright.",
"Hana is bright.",
"Evan is large.",
"Dalia is calm.",
"Evan is quiet.",
"Dalia is quiet.",
"Farah is bright.",
"Hana is large."
] | [
"For every person mentioned: if person 1 is calm, then person 1 is kind.",
"For every person mentioned: if person 1 is kind, then person 1 is warm.",
"For every person mentioned: if person 1 is not calm, then person 1 is not kind.",
"For every person mentioned: if person 1 is not kind, then person 1 is not wa... | Evan is not warm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is large.
Dalia is large.
Hana is calm.
Farah is large.
Gabe is quiet.
Farah is qui... |
v7-r1-calibration-old-000128-calibration-true | v7-r1-calibration-old-000128 | calibration | {
"entities": [
"EntityA",
"Dalia",
"Hana",
"Chen"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": true
},
{
"predicate": "blue",
"arguments": [
"Hana"
],
"negative": false
},
{
"pr... | {
"predicate": "quiet",
"arguments": [
"EntityA"
],
"negative": true
} | TRUE | 1 | 1 | chain | canonical | 8d934dddd98394eb66c433e6e730ee187a3d597fd8d524c7f6b9f1217c6051a0 | d6cdea0617b6466794837beafb7653bbf07acb9494a5c628e1b8f13f481fdb99 | [
"Dalia is not blue.",
"EntityA is large.",
"Dalia is not round.",
"Dalia is warm.",
"Hana is not bright.",
"Hana is blue.",
"Chen is not warm.",
"Hana is kind.",
"EntityA is not calm."
] | [
"For every person mentioned: if person 1 is large and person 1 is calm, then person 1 is quiet.",
"For every person mentioned: if person 1 is large and person 1 is not calm, then person 1 is not quiet."
] | EntityA is not quiet. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is not blue.
EntityA is large.
Dalia is not round.
Dalia is warm.
Hana is not brig... |
v7-r1-calibration-multi-5-0004-2-multi-base-1 | v7-r1-calibration-multi-5-0004-2 | calibration | {
"entities": [
"Evan",
"Farah",
"Gabe",
"Hana",
"EntityA"
],
"facts": [
{
"predicate": "quiet",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "large",
"arguments": [
"Gabe"
],
"negative": false
}... | {
"predicate": "calm",
"arguments": [
"Farah"
],
"negative": false
} | UNKNOWN | null | 2 | conjunction | canonical | 184374441dacf1da25666b414fbc9da0e7252cdc6d1639fa42acc976f094ae9f | b9e4d0d1df2e155485fe046c7c8ffccaeac8b4300ebcd3a6c2c416f934842044 | [
"Hana is round.",
"Gabe is round.",
"EntityA is not large.",
"Gabe is bright.",
"EntityA is round.",
"Hana is bright.",
"Gabe is large.",
"Farah is round.",
"Farah is quiet.",
"Evan is round.",
"Evan is quiet.",
"Farah is bright.",
"Evan is bright.",
"Hana is quiet.",
"EntityA is quiet."... | [
"For every person mentioned: if person 1 is not blue, then person 1 is not calm.",
"For every person mentioned: if person 1 is round and person 1 is large, then person 1 is blue.",
"For every person mentioned: if person 1 is round and person 1 is not large, then person 1 is not blue.",
"For every person menti... | Farah is calm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Hana is round.
Gabe is round.
EntityA is not large.
Gabe is bright.
EntityA is round.
Ha... |
v7-r1-calibration-old-000135-calibration-true | v7-r1-calibration-old-000135 | calibration | {
"entities": [
"Chen",
"EntityA",
"Farah",
"Hana"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Chen"
],
"negative": true
},
{
"predicate": "kind",
"arguments": [
"EntityA"
],
"negative": false
},
{
"... | {
"predicate": "quiet",
"arguments": [
"Chen"
],
"negative": true
} | TRUE | 1 | 1 | chain | canonical | c5f2b15743ae37bc15f563f4cdd2638b6ec51e369e0aa2c9ca29bba3ccfaf39b | d6cdea0617b6466794837beafb7653bbf07acb9494a5c628e1b8f13f481fdb99 | [
"Chen is not blue.",
"EntityA is kind.",
"Farah is not round.",
"EntityA is warm.",
"Hana is not kind.",
"Chen is large.",
"Hana is warm.",
"EntityA is not quiet."
] | [
"For every person mentioned: if person 1 is not blue and person 1 is not large, then person 1 is quiet.",
"For every person mentioned: if person 1 is not blue and person 1 is large, then person 1 is not quiet."
] | Chen is not quiet. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Chen is not blue.
EntityA is kind.
Farah is not round.
EntityA is warm.
Hana is not kind... |
v7-r1-calibration-old-000037-calibration-false | v7-r1-calibration-old-000037 | calibration | {
"entities": [
"Evan",
"EntityA",
"Hana",
"Dalia"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"EntityA"
],
"negative": false
},
{
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": true
},
{
... | {
"predicate": "bright",
"arguments": [
"Evan"
],
"negative": false
} | FALSE | 1 | 1 | chain | conditional | 6918542d3b9b8893e7dd58563707e80942086eb929d04f8d4c298cb32b5ed30b | d6cdea0617b6466794837beafb7653bbf07acb9494a5c628e1b8f13f481fdb99 | [
"Dalia is warm.",
"Evan is not large.",
"Dalia is not bright.",
"EntityA is blue.",
"EntityA is bright.",
"Hana is kind.",
"Evan is not round.",
"Dalia is not blue."
] | [
"Whenever person 1 is not large and person 1 is not round, person 1 is not bright; this applies to all people mentioned.",
"Whenever person 1 is not large and person 1 is round, person 1 is bright; this applies to all people mentioned."
] | Evan is bright. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is warm.
Evan is not large.
Dalia is not bright.
EntityA is blue.
EntityA is brigh... |
v7-r1-calibration-old-000123-calibration-true | v7-r1-calibration-old-000123 | calibration | {
"entities": [
"Dalia",
"Bella",
"Gabe"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Dalia"
],
"negative": true
},
{
"predicate": "approved",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate... | {
"predicate": "verified",
"arguments": [
"Bella"
],
"negative": true
} | TRUE | 2 | 2 | relational | canonical | 9c379752a66782b8ed827cb67e2691da2cf30c3577351e3306307a945796b35a | 44683fac69fbc06ccbf74f16e30207e66c2ec7d4a3a61270b430a85a65b794c8 | [
"Dalia is not active.",
"Bella follows Dalia.",
"Gabe is not verified.",
"Gabe is ready.",
"Dalia is approved.",
"Bella is enabled.",
"Dalia is not ready."
] | [
"For every person mentioned: if person 1 is trusted and person 1 is enabled, then person 1 is not verified.",
"For every person mentioned: if person 1 follows person 2 and person 2 is not active, then person 1 is trusted.",
"For every person mentioned: if person 1 is trusted and person 1 is not enabled, then pe... | Bella is not verified. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Dalia is not active.
Bella follows Dalia.
Gabe is not verified.
Gabe is ready.
Dalia is ... |
v7-r1-calibration-old-000112-calibration-unknown | v7-r1-calibration-old-000112 | calibration | {
"entities": [
"Bella",
"Gabe",
"Hana"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Bella"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Gabe"
],
"negative": false
},
{
"predicate": "br... | {
"predicate": "warm",
"arguments": [
"Gabe"
],
"negative": true
} | UNKNOWN | null | 1 | chain | canonical | 93b12cd0158cd01c7ad5807ff650348d9610fc4cb89f1cd7c1f943f409449ff3 | 452fc03fe2b109b042d10502e5770d53dc7985bb632c4f2fa96a4e6b8a8a36dc | [
"Gabe is not quiet.",
"Gabe is large.",
"Bella is not round.",
"Hana is calm.",
"Hana is bright.",
"Hana is not warm.",
"Bella is not blue.",
"Gabe is bright.",
"Hana is not large."
] | [
"For every person mentioned: if person 1 is not blue and person 1 is round, then person 1 is warm.",
"For every person mentioned: if person 1 is not blue and person 1 is not round, then person 1 is not warm."
] | Gabe is not warm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is not quiet.
Gabe is large.
Bella is not round.
Hana is calm.
Hana is bright.
Hana... |
v7-r1-calibration-multi-5-0003-1-multi-base-0 | v7-r1-calibration-multi-5-0003-1 | calibration | {
"entities": [
"Dalia",
"Evan",
"Farah",
"Gabe",
"Hana"
],
"facts": [
{
"predicate": "enabled",
"arguments": [
"Evan"
],
"negative": false
},
{
"predicate": "enabled",
"arguments": [
"Farah"
],
"negative": false
}... | {
"predicate": "active",
"arguments": [
"Dalia"
],
"negative": true
} | TRUE | 1 | 1 | signed_chain | canonical | af4435539dd6b46b430e01f123012dbcb0a411f0996c8bea37aeef2322eb8c9a | 305f008cc3b4bb258b4caa98d4f7273ffc21d1ca43c606af6f187003df0bde3d | [
"Gabe is approved.",
"Dalia is enabled.",
"Gabe is enabled.",
"Evan is not authorized.",
"Evan is approved.",
"Gabe is authorized.",
"Hana is enabled.",
"Dalia is not authorized.",
"Farah is approved.",
"Hana is approved.",
"Evan is enabled.",
"Dalia is trusted.",
"Farah is enabled.",
"Far... | [
"For every person mentioned: if person 1 is not authorized, then person 1 is not active.",
"For every person mentioned: if person 1 is authorized, then person 1 is active."
] | Dalia is not active. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Gabe is approved.
Dalia is enabled.
Gabe is enabled.
Evan is not authorized.
Evan is app... |
v7-r1-calibration-multi-4-0005-1-multi-reordered-1 | v7-r1-calibration-multi-4-0005-1 | calibration | {
"entities": [
"Farah",
"Gabe",
"Hana",
"EntityA"
],
"facts": [
{
"predicate": "verified",
"arguments": [
"Farah"
],
"negative": false
},
{
"predicate": "registered",
"arguments": [
"Farah"
],
"negative": false
},
... | {
"predicate": "trusted",
"arguments": [
"Gabe"
],
"negative": true
} | FALSE | 2 | 2 | conjunction | canonical | 6e0b364ec2c6d6921eeb7b33450a9b7b6a6dfe0f28baf1887cf7da2f7ebb1166 | 706cae65cedc1b3d33cb62d3f2a295e30222a06f05857cad6e0a0711fd327dd7 | [
"EntityA is verified.",
"EntityA is enabled.",
"Hana is verified.",
"Gabe is ready.",
"Gabe is registered.",
"EntityA is ready.",
"Farah is registered.",
"Hana is not ready.",
"Hana is enabled.",
"Gabe is verified.",
"Farah is enabled.",
"Farah is verified."
] | [
"For every person mentioned: if person 1 is verified and person 1 is not ready, then person 1 is not active.",
"For every person mentioned: if person 1 is active, then person 1 is trusted.",
"For every person mentioned: if person 1 is verified and person 1 is ready, then person 1 is active.",
"For every perso... | Gabe is not trusted. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
EntityA is verified.
EntityA is enabled.
Hana is verified.
Gabe is ready.
Gabe is regist... |
v7-r1-calibration-old-000110-calibration-unknown | v7-r1-calibration-old-000110 | calibration | {
"entities": [
"Farah",
"Dalia",
"Bella",
"Hana"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Dalia"
],
"negative": true
},
{
"predicate": "approved",
"arguments": [
"Farah"
],
"negative": true
},
{
... | {
"predicate": "enabled",
"arguments": [
"Dalia"
],
"negative": true
} | UNKNOWN | null | 2 | chain | conditional | aa0950c9bb7e784bb3dd5b9c2ff0af150573c94f90c07be7ccd67b8cfa6046fd | 452fc03fe2b109b042d10502e5770d53dc7985bb632c4f2fa96a4e6b8a8a36dc | [
"Hana is verified.",
"Hana is enabled.",
"Dalia is authorized.",
"Farah is trusted.",
"Dalia is not active.",
"Hana is registered.",
"Farah is not approved."
] | [
"Whenever person 1 is verified and person 1 is not approved, person 1 is enabled; this applies to all people mentioned.",
"Whenever person 1 is verified and person 1 is approved, person 1 is not enabled; this applies to all people mentioned.",
"Whenever person 1 is trusted, person 1 is verified; this applies to... | Dalia is not enabled. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Hana is verified.
Hana is enabled.
Dalia is authorized.
Farah is trusted.
Dalia is not a... |
v7-r1-calibration-old-000101-calibration-true | v7-r1-calibration-old-000101 | calibration | {
"entities": [
"EntityA",
"Chen",
"Hana",
"Dalia"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Dalia"
],
"negative": false
},
{
"predicate": "blue",
"arguments": [
"Hana"
],
"negative": true
},
{
"pr... | {
"predicate": "kind",
"arguments": [
"EntityA"
],
"negative": false
} | TRUE | 1 | 1 | chain | conditional | 3142bf9d8153f60005ed507f6b1855ae5178fd2acb9d7ae1c5312d668b746e56 | d6cdea0617b6466794837beafb7653bbf07acb9494a5c628e1b8f13f481fdb99 | [
"Hana is not blue.",
"Dalia is not round.",
"Dalia is not calm.",
"EntityA is not quiet.",
"Chen is large.",
"EntityA is warm.",
"Dalia is blue.",
"Hana is round."
] | [
"Whenever person 1 is not quiet and person 1 is not warm, person 1 is not kind; this applies to all people mentioned.",
"Whenever person 1 is not quiet and person 1 is warm, person 1 is kind; this applies to all people mentioned."
] | EntityA is kind. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Hana is not blue.
Dalia is not round.
Dalia is not calm.
EntityA is not quiet.
Chen is l... |
v7-r1-calibration-old-000125-calibration-false | v7-r1-calibration-old-000125 | calibration | {
"entities": [
"Bella",
"Farah",
"Evan"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Farah"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Evan"
],
"negative": false
},
{
"predicate": "c... | {
"predicate": "calm",
"arguments": [
"Bella"
],
"negative": false
} | FALSE | 3 | 3 | chain | canonical | 90c47a4bc35198be8933949a446e06666f91e30fa11a9b4fc5c3c0aa1661ce34 | 8628c5aad256c8edba6f7f2657c0483db19285b433a26e9bbaccd959071fc344 | [
"Farah is calm.",
"Evan is bright.",
"Farah is not blue.",
"Bella is not kind.",
"Bella is not round.",
"Evan is quiet."
] | [
"For every person mentioned: if person 1 is not kind, then person 1 is warm.",
"For every person mentioned: if person 1 is quiet and person 1 is not round, then person 1 is not calm.",
"For every person mentioned: if person 1 is quiet and person 1 is round, then person 1 is calm.",
"For every person mentioned... | Bella is calm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Farah is calm.
Evan is bright.
Farah is not blue.
Bella is not kind.
Bella is not round.... |
v7-r1-calibration-old-000020-calibration-true | v7-r1-calibration-old-000020 | calibration | {
"entities": [
"Evan",
"Gabe",
"EntityA",
"Hana"
],
"facts": [
{
"predicate": "active",
"arguments": [
"Evan"
],
"negative": true
},
{
"predicate": "approved",
"arguments": [
"Gabe"
],
"negative": false
},
{
... | {
"predicate": "verified",
"arguments": [
"Gabe"
],
"negative": true
} | TRUE | 2 | 2 | relational | canonical | 8e51ae7684dac8622bc8b8142d6efb1e9e6cb2ddf8a90d256f5cd5b6fa95b961 | 44683fac69fbc06ccbf74f16e30207e66c2ec7d4a3a61270b430a85a65b794c8 | [
"Evan is not active.",
"EntityA is registered.",
"Gabe follows Evan.",
"EntityA is not enabled.",
"Hana is trusted.",
"Evan is registered.",
"Gabe is approved.",
"EntityA is not trusted."
] | [
"For every person mentioned: if person 1 is trusted and person 1 is approved, then person 1 is not verified.",
"For every person mentioned: if person 1 follows person 2 and person 2 is not active, then person 1 is trusted.",
"For every person mentioned: if person 1 is trusted and person 1 is not approved, then ... | Gabe is not verified. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Evan is not active.
EntityA is registered.
Gabe follows Evan.
EntityA is not enabled.
Ha... |
v7-r1-calibration-old-000036-calibration-true | v7-r1-calibration-old-000036 | calibration | {
"entities": [
"Evan",
"Gabe",
"Farah",
"Hana"
],
"facts": [
{
"predicate": "blue",
"arguments": [
"Gabe"
],
"negative": true
},
{
"predicate": "bright",
"arguments": [
"Evan"
],
"negative": false
},
{
"pred... | {
"predicate": "calm",
"arguments": [
"Evan"
],
"negative": true
} | TRUE | 1 | 1 | relational | canonical | 15e1266d4571d5e902a185f17a9ca28c471a388b98c368f9364c930f5205a602 | 5b711a9c568ac2fd187eae882b1a35a74b5549b53da284a9e55be83fad8e4b23 | [
"Farah is not round.",
"Farah is not calm.",
"Evan is not large.",
"Evan is bright.",
"Hana is not round.",
"Gabe is kind.",
"Farah is not kind.",
"Gabe is not blue.",
"Gabe is not quiet."
] | [
"For every person mentioned: if person 1 is not large and person 1 is bright, then person 1 is not calm.",
"For every person mentioned: if person 1 is not large and person 1 is not bright, then person 1 is calm."
] | Evan is not calm. | Decide using only these facts and rules. TRUE means the statement follows; FALSE means its explicit negation follows; UNKNOWN means neither follows. Absence is not negation. Answer A for TRUE, B for FALSE, or C for UNKNOWN.
Facts:
Farah is not round.
Farah is not calm.
Evan is not large.
Evan is bright.
Hana is not ro... |
End of preview. Expand in Data Studio
decipop worlds
Synthetic facts and rules for testing whether a small model follows changing evidence. Each question has an exact TRUE, FALSE or UNKNOWN answer from a symbolic oracle.
Load
from datasets import load_dataset
rows = load_dataset("axondendriteplus/decipop-worlds", "v8-r2", split="test_paired")
print(rows[0]["prompt"], rows[0]["label"])
Choose v7-r1, v7-r2, v8-r1 or v8-r2. The release contains 190,688 serialized
rows, including repeated training questions and related transformations. They are
not independent worlds. See row-counts.json for every split count.
What changed in this release
This is a name-normalized derivative of the frozen V7/V8 research data. One entity
symbol is replaced with EntityA throughout symbolic theories, queries and rendered
text. Labels, proof depths, IDs, split membership and family relationships are
preserved. Every question was checked with two symbolic oracles; sampled canonical
world identities were also checked. normalization.json records the checks, and
sha256.json records both source and derived file hashes. Original manifests are
not presented as checksums for these changed files.
The published model scores use the original input text. They are not scores on
this renamed release. Renaming preserves logical meaning but can change model
tokenization and predictions. The released checkpoints have not been re-evaluated
on the normalized inputs. Historical source and results remain at the code's
v0.1.0 tag; the original dataset is retained privately for exact reproduction.
Task and schema
TRUE: the queried statement follows. FALSE: its explicit negation follows. UNKNOWN: neither follows. Missing evidence is not negative evidence. UNKNOWN is not a low-confidence refusal. Contradictory worlds are rejected.
Rows contain example_id, family_id, split, theory, query, label,
proof_depth, theory_depth, canonical hashes, motif, renderer_id, rendered
facts, rules, question and prompt. The prompt's A/B/C choices map to
TRUE/FALSE/UNKNOWN. Names are arbitrary generated symbols. Family registries live
under provenance/. No teacher-model labels or customer data are included.
Splits and overlap
train_control and train_treatment are experimental arms, not disjoint splits.
V8's training files are identical by design. Development, calibration, selection-fit
and selection use distinct families within a config. Related queries and edits
share a family and must stay together. Repetitions exclude prior canonical worlds,
but templates, vocabulary and rule skeletons are shared. Later studies respond to
earlier findings: do not merge configs into a single uncontaminated benchmark.
Question accuracy counts individual rows. Complete-world accuracy requires every entity correct. Complete paired worlds additionally require every query correct after an assigned edit. Complete transformation families require all included variants. UNKNOWN edits can be identity controls; inspect known-target edits separately. The test labels are public; tuning on them forfeits a fresh test claim.
Intended use and limits
For controlled research on entity tracking, evidence sensitivity, calibration and checkpoint selection. This narrow synthetic distribution does not establish reliable business decisions, unrestricted reasoning or real-world safety.
Project-generated data: Apache 2.0. Related work includes RuleTaker and ProofWriter; their data are not redistributed here. Model-base attribution is separate.
- Downloads last month
- 80