triple1 listlengths 3 3 | triple2 listlengths 3 3 | triple3 listlengths 3 3 | triple1_labels listlengths 3 3 | triple2_labels listlengths 3 3 | triple3_labels listlengths 3 3 | description stringlengths 10 16 | keep_triple bool 1
class | query_prompt stringlengths 3.24k 3.31k | query stringlengths 90 171 | entity_a stringlengths 6 29 | entity_b stringlengths 5 37 | rel_b stringlengths 4 22 | prompt stringlengths 3.56k 3.64k | index int64 420 439 | raw_responses stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
[
"Q1291701",
"P463",
"Q11895"
] | [
"Q11885",
"P463",
"Q11895"
] | [
"Q11885",
"P1303",
"Q17172850"
] | [
"Steven Adler",
"member of",
"Guns N' Roses"
] | [
"Axl Rose",
"member of",
"Guns N' Roses"
] | [
"Axl Rose",
"instrument",
"voice"
] | Q11895__P463 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Steven Adler, the drummer known for his work with Guns N' Roses, and someone who is recognized for their vocal abilities in the band? | Steven Adler | voice | instrument | Query: What are different ways of connecting Steven Adler, the drummer known for his work with Guns N' Roses, and someone who is recognized for their vocal abilities in the band?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual trip... | 420 | {} |
[
"Q104898827",
"P162",
"Q95468740"
] | [
"Q104898827",
"P162",
"Q115728320"
] | [
"Q115728320",
"P106",
"Q1049296"
] | [
"Shadowplay",
"producer",
"Vendula Bradáčová"
] | [
"Shadowplay",
"producer",
"Maja Hamplová"
] | [
"Maja Hamplová",
"occupation",
"casting director"
] | Q104898827__P162 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Vendula Bradáčová and someone who works as a casting director? | Vendula Bradáčová | casting director | occupation | Query: What are different ways of connecting Vendula Bradáčová and someone who works as a casting director?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a continuous chain that begins w... | 421 | {} |
[
"Q229836",
"P40",
"Q1411"
] | [
"Q156778",
"P40",
"Q1411"
] | [
"Q156778",
"P5025",
"Q385338"
] | [
"Antonia Minor",
"child",
"Claudius"
] | [
"Nero Claudius Drusus",
"child",
"Claudius"
] | [
"Nero Claudius Drusus",
"gens",
"Claudia gens"
] | Q1411__P40 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Antonia Minor, a prominent figure in Roman history, and someone who is associated with the Claudia gens? | Antonia Minor | Claudia gens | gens | Query: What are different ways of connecting Antonia Minor, a prominent figure in Roman history, and someone who is associated with the Claudia gens?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail... | 422 | {} |
[
"Q101091534",
"P162",
"Q101090806"
] | [
"Q101091534",
"P162",
"Q101090684"
] | [
"Q101090684",
"P106",
"Q3387717"
] | [
"The Euphoria of Being",
"producer",
"Marcell Gerő"
] | [
"The Euphoria of Being",
"producer",
"Réka Szabó"
] | [
"Réka Szabó",
"occupation",
"theatrical director"
] | Q101091534__P162 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Marcell Gerő, a notable figure in the arts, and someone who works as a theatrical director? | Marcell Gerő | theatrical director | occupation | Query: What are different ways of connecting Marcell Gerő, a notable figure in the arts, and someone who works as a theatrical director?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a c... | 423 | {} |
[
"Q725726",
"P54",
"Q188984"
] | [
"Q2094838",
"P54",
"Q188984"
] | [
"Q2094838",
"P1532",
"Q30"
] | [
"Sean Avery",
"member of sports team",
"New York Rangers"
] | [
"Ryan McDonagh",
"member of sports team",
"New York Rangers"
] | [
"Ryan McDonagh",
"country for sport",
"United States of America"
] | Q188984__P54 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Sean Avery, the former professional ice hockey player, and someone who represents the United States of America in their sport? | Sean Avery | United States of America | country for sport | Query: What are different ways of connecting Sean Avery, the former professional ice hockey player, and someone who represents the United States of America in their sport?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (he... | 424 | {} |
[
"Q83071",
"P161",
"Q16203452"
] | [
"Q83071",
"P161",
"Q8062292"
] | [
"Q8062292",
"P106",
"Q10800557"
] | [
"Downton Abbey",
"cast member",
"Stephen Critchlow"
] | [
"Downton Abbey",
"cast member",
"Yves Aubert"
] | [
"Yves Aubert",
"occupation",
"film actor"
] | Q83071__P161 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Stephen Critchlow, a talented voice actor, and someone who is known for their work as a film actor? | Stephen Critchlow | film actor | occupation | Query: What are different ways of connecting Stephen Critchlow, a talented voice actor, and someone who is known for their work as a film actor?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) for... | 425 | {} |
[
"Q18041364",
"P2293",
"Q126691"
] | [
"Q18035557",
"P2293",
"Q126691"
] | [
"Q18035557",
"P684",
"Q18301613"
] | [
"CDKAL1",
"genetic association",
"gestational diabetes"
] | [
"IGF2BP2",
"genetic association",
"gestational diabetes"
] | [
"IGF2BP2",
"ortholog",
"Igf2bp2"
] | Q126691__P2293 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting CDKAL1, a gene associated with diabetes, and a substance that is an ortholog of IGF2BP2? | CDKAL1 | Igf2bp2 | ortholog | Query: What are different ways of connecting CDKAL1, a gene associated with diabetes, and a substance that is an ortholog of IGF2BP2?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a cont... | 426 | {} |
[
"Q33524365",
"P463",
"Q18557"
] | [
"Q2706536",
"P463",
"Q18557"
] | [
"Q2706536",
"P26",
"Q3336630"
] | [
"Waltteri Väyrynen",
"member of",
"Opeth"
] | [
"Martin Axenrot",
"member of",
"Opeth"
] | [
"Martin Axenrot",
"spouse",
"Nathalie Lorichs"
] | Q18557__P463 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Waltteri Väyrynen, the Finnish drummer, and someone who is the spouse of Nathalie Lorichs? | Waltteri Väyrynen | Nathalie Lorichs | spouse | Query: What are different ways of connecting Waltteri Väyrynen, the Finnish drummer, and someone who is the spouse of Nathalie Lorichs?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a co... | 427 | {} |
[
"Q229413",
"P40",
"Q1409"
] | [
"Q191039",
"P40",
"Q1409"
] | [
"Q191039",
"P3716",
"Q98108963"
] | [
"Agrippina the Elder",
"child",
"Caligula"
] | [
"Germanicus",
"child",
"Caligula"
] | [
"Germanicus",
"social classification",
"patricius"
] | Q1409__P40 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Agrippina the Elder, a prominent figure in ancient Rome, and someone who is classified as a patrician? | Agrippina the Elder | patricius | social classification | Query: What are different ways of connecting Agrippina the Elder, a prominent figure in ancient Rome, and someone who is classified as a patrician?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) ... | 428 | {} |
[
"Q571032",
"P161",
"Q516690"
] | [
"Q571032",
"P161",
"Q3875987"
] | [
"Q3875987",
"P27",
"Q804"
] | [
"Django Unchained",
"cast member",
"Walton Goggins"
] | [
"Django Unchained",
"cast member",
"Nichole Galicia"
] | [
"Nichole Galicia",
"country of citizenship",
"Panama"
] | Q571032__P161 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Walton Goggins, the American actor, and someone who is a citizen of Panama? | Walton Goggins | Panama | country of citizenship | Query: What are different ways of connecting Walton Goggins, the American actor, and someone who is a citizen of Panama?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a continuous chain ... | 429 | {} |
[
"Q1004567",
"P162",
"Q722890"
] | [
"Q1004567",
"P162",
"Q542003"
] | [
"Q542003",
"P19",
"Q151920"
] | [
"Bullseye!",
"producer",
"Michael Winner"
] | [
"Bullseye!",
"producer",
"Menahem Golan"
] | [
"Menahem Golan",
"place of birth",
"Tiberias"
] | Q1004567__P162 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Michael Winner, the British film director, and someone who was born in Tiberias? | Michael Winner | Tiberias | place of birth | Query: What are different ways of connecting Michael Winner, the British film director, and someone who was born in Tiberias?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a continuous c... | 430 | {} |
[
"Q59267910",
"P161",
"Q47300912"
] | [
"Q241681",
"P161",
"Q47300912"
] | [
"Q47300912",
"P19",
"Q1144366"
] | [
"Rachel Redleaf",
"cast member",
"Once Upon a Time in Hollywood"
] | [
"Rebecca Gayheart",
"cast member",
"Once Upon a Time in Hollywood"
] | [
"Once Upon a Time in Hollywood",
"place of birth",
"Hazard"
] | Q47300912__P161 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting "Once Upon a Time in Hollywood," the film, and someone who was born in a place associated with the film's title? | Once Upon a Time in Hollywood | Hazard | place of birth | Query: What are different ways of connecting "Once Upon a Time in Hollywood," the film, and someone who was born in a place associated with the film's title?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationsh... | 431 | {} |
[
"Q320065",
"P166",
"Q935843"
] | [
"Q162667",
"P166",
"Q935843"
] | [
"Q162667",
"P106",
"Q33999"
] | [
"Artie Shaw",
"award received",
"Grammy Lifetime Achievement Award"
] | [
"Glen Campbell",
"award received",
"Grammy Lifetime Achievement Award"
] | [
"Glen Campbell",
"occupation",
"actor"
] | Q935843__P166 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Artie Shaw, the famous clarinetist and bandleader, and someone who is known for their work as an actor? | Artie Shaw | actor | occupation | Query: What are different ways of connecting Artie Shaw, the famous clarinetist and bandleader, and someone who is known for their work as an actor?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail)... | 432 | {} |
[
"Q412236",
"P2175",
"Q12192"
] | [
"Q420595",
"P2175",
"Q12192"
] | [
"Q420595",
"P2868",
"Q804539"
] | [
"voriconazole",
"medical condition treated",
"pneumonia"
] | [
"tigecycline",
"medical condition treated",
"pneumonia"
] | [
"tigecycline",
"subject has role",
"bactericide"
] | Q12192__P2175 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting voriconazole, an antifungal medication, and a substance that has a role as a bactericide? | voriconazole | bactericide | subject has role | Query: What are different ways of connecting voriconazole, an antifungal medication, and a substance that has a role as a bactericide?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a con... | 433 | {} |
[
"Q557148",
"P54",
"Q169138"
] | [
"Q2472334",
"P54",
"Q169138"
] | [
"Q2472334",
"P1532",
"Q215"
] | [
"P. J. Brown",
"member of sports team",
"Miami Heat"
] | [
"Zoran Dragić",
"member of sports team",
"Miami Heat"
] | [
"Zoran Dragić",
"country for sport",
"Slovenia"
] | Q169138__P54 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting P. J. Brown, a notable figure in basketball, and someone who represents Slovenia in the sport? | P. J. Brown | Slovenia | country for sport | Query: What are different ways of connecting P. J. Brown, a notable figure in basketball, and someone who represents Slovenia in the sport?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming ... | 434 | {} |
[
"Q167520",
"P40",
"Q13909"
] | [
"Q235845",
"P40",
"Q13909"
] | [
"Q235845",
"P451",
"Q1701970"
] | [
"Jon Voight",
"child",
"Angelina Jolie"
] | [
"Marcheline Bertrand",
"child",
"Angelina Jolie"
] | [
"Marcheline Bertrand",
"unmarried partner",
"John Trudell"
] | Q13909__P40 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Jon Voight, the American actor, and someone who was an unmarried partner of John Trudell? | Jon Voight | John Trudell | unmarried partner | Query: What are different ways of connecting Jon Voight, the American actor, and someone who was an unmarried partner of John Trudell?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a con... | 435 | {} |
[
"Q1024838",
"P58",
"Q15430128"
] | [
"Q1024838",
"P58",
"Q358714"
] | [
"Q358714",
"P19",
"Q49196"
] | [
"Cabin Fever",
"screenwriter",
"Randy Pearlstein"
] | [
"Cabin Fever",
"screenwriter",
"Eli Roth"
] | [
"Eli Roth",
"place of birth",
"Newton"
] | Q1024838__P58 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting Randy Pearlstein and someone who was born in Newton? | Randy Pearlstein | Newton | place of birth | Query: What are different ways of connecting Randy Pearlstein and someone who was born in Newton?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a continuous chain that begins with one en... | 436 | {} |
[
"Q18043072",
"P2293",
"Q11081"
] | [
"Q18046351",
"P2293",
"Q11081"
] | [
"Q18046351",
"P5572",
"Q66558397"
] | [
"PLEKHG1",
"genetic association",
"Alzheimer's disease"
] | [
"CLMN",
"genetic association",
"Alzheimer's disease"
] | [
"CLMN",
"expressed in",
"lateral nuclear group of thalamus"
] | Q11081__P2293 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting PLEKHG1, a gene, and a substance that is expressed in the lateral nuclear group of the thalamus? | PLEKHG1 | lateral nuclear group of thalamus | expressed in | Query: What are different ways of connecting PLEKHG1, a gene, and a substance that is expressed in the lateral nuclear group of the thalamus?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) formin... | 437 | {} |
[
"Q18041790",
"P2293",
"Q11088"
] | [
"Q18027568",
"P2293",
"Q11088"
] | [
"Q18027568",
"P688",
"Q21139492"
] | [
"STK32B",
"genetic association",
"coeliac disease"
] | [
"IFI16",
"genetic association",
"coeliac disease"
] | [
"IFI16",
"encodes",
"Gamma-interferon-inducible protein 16"
] | Q11088__P2293 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting STK32B, a gene, and a substance that encodes the Gamma-interferon-inducible protein 16? | STK32B | Gamma-interferon-inducible protein 16 | encodes | Query: What are different ways of connecting STK32B, a gene, and a substance that encodes the Gamma-interferon-inducible protein 16?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relationship, tail) forming a conti... | 438 | {} |
[
"Q10305218",
"P57",
"Q2470131"
] | [
"Q10305218",
"P57",
"Q10311645"
] | [
"Q10311645",
"P800",
"Q27703588"
] | [
"Window of the Soul",
"director",
"Walter Carvalho"
] | [
"Window of the Soul",
"director",
"João Jardim"
] | [
"João Jardim",
"notable work",
"Pro Dia Nascer Feliz"
] | Q10305218__P57 | false | Task: Generate a Natural Language Query.
You will be given input in the following format:
Concept A: entity1
Concept B: entity2
Concept C: (concept B, rel, entity3)
Concept A and Concept C are connected via an independent concept B. Concept C is a property of Concept B, which the person is supposed to guess.
You... | What are different ways of connecting the concept of the "Window of the Soul" and someone who is known for their notable work on "Pro Dia Nascer Feliz"? | Window of the Soul | Pro Dia Nascer Feliz | notable work | Query: What are different ways of connecting the concept of the "Window of the Soul" and someone who is known for their notable work on "Pro Dia Nascer Feliz"?
Task: Identify how two real-world entities are connected by producing MANY connection paths. A connection path is a sequence of factual triples (head, relation... | 439 | {} |
Dataset: connections-dev/connection_queries_jan12
This dataset was generated using the inference script with the following configuration:
Inference Parameters
Model Configuration
- Model Name:
gemini-3-pro-preview - Server URL:
Not specified - API Key:
Not provided - Request Timeout:
30 seconds
Query Configuration
- Query Type:
natural - Query Column:
query - Sampling Type:
original
Generation Parameters
- Temperature:
0.7 - Max Tokens:
16384 - Number of Runs:
1 - Chunk Size:
Entire dataset - Effort Level:
high
Feature Flags
- Step by Step Reasoning:
False - Reasoning Enabled:
False - No Curator:
False - Index Filter:
None (all samples) - Start Index:
420 - End Index:
440
Data Information
- Input File:
connections-dev/connection_queries_jan12 - Number of Samples:
20 - Output Filename:
connection_queries_jan12__natural__s420__e440__original__1__high__0.7__16384__gemini-3-pro-preview.jsonl
Dataset Structure
The dataset contains the following key columns:
query: The original queryentity_a: Starting entityentity_b: Target entityrel_b: Target relationshippath_prediction: Generated connection pathsprompt_for_path: Prompts used for path generation
Usage
You can load this dataset using:
from datasets import load_dataset
dataset = load_dataset("s420_e440_original_1_high_0.7_16384_gemini-3-pro-preview__raw")
Generation Details
This dataset was generated on 2026-02-10 14:23:42 using the inference pipeline with the above configuration.
- Downloads last month
- 6