Hypencoder: Hypernetworks for Information Retrieval
Paper • 2502.05364 • Published • 10
query dict | items listlengths 201 207 |
|---|---|
{
"content": "define extreme",
"id": "121352",
"tokenized_content": [
101,
9375,
6034,
102
]
} | [
{
"content": "Definition of extreme. 1 1a : something situated at or marking one end or the other of a range extremes of heat and coldb : the first term or the last term of a mathematical proportionc : the major term or minor term of a syllogism. 2 2a : a very pronounced or excessive degreeb : highest degree ... |
{"content":"what does chattel mean on credit history","id":"634306","tokenized_content":[101,2054,25(...TRUNCATED) | [{"content":"commercial tort claims, deposit accounts, lette rs of credit, or rights arising out of (...TRUNCATED) |
{"content":"what was the great leap forward brainly","id":"920825","tokenized_content":[101,2054,200(...TRUNCATED) | [{"content":"The Great Leap Forward (Chinese: 大跃进; pinyin: Dà yuè jìn) of the People's Repu(...TRUNCATED) |
{"content":"tattoo fixers how much does it cost","id":"510633","tokenized_content":[101,11660,8081,2(...TRUNCATED) | [{"content":"1 Laser tattoo removal, which is the most common method, and usually the best, ranges f(...TRUNCATED) |
{"content":"what is decentralization process.","id":"737889","tokenized_content":[101,2054,2003,1151(...TRUNCATED) | [{"content":"Decentralization is the process of redistributing or dispersing functions, powers, peop(...TRUNCATED) |
{"content":"what is a bank transit number","id":"674172","tokenized_content":[101,2054,2003,1037,292(...TRUNCATED) | [{"content":"A bank's Routing Transit Number (RTN) is a 9-digit code used to identify your bank on c(...TRUNCATED) |
{"content":"how much can i contribute to nondeductible ira","id":"303205","tokenized_content":[101,2(...TRUNCATED) | [{"content":"Nondeductible IRA Contributions. For a traditional IRA, you're allowed to contribute up(...TRUNCATED) |
{"content":"what are the four major groups of elements","id":"570009","tokenized_content":[101,2054,(...TRUNCATED) | [{"content":"The elements in the periodic table are often divided into four categories: (1) main gro(...TRUNCATED) |
{
"content": "sanitizer temperature",
"id": "492875",
"tokenized_content": [
101,
2624,
25090,
6290,
4860,
102
]
} | [{"content":"Chlorine sanitizing solutions should be at a minimum temperature of 75°F. They are les(...TRUNCATED) |
{"content":"blood clots in urine after menopause","id":"54528","tokenized_content":[101,2668,18856,1(...TRUNCATED) | [{"content":"1 Blood clots AND Vaginal bleeding after menopause: Causes of All Symptoms. 2 Blood cl(...TRUNCATED) |
End of preview. Expand in Data Studio
The MSMARCO training data used to train the models from Hypencoder: Hypernetworks for Information Retrieval .
Dataset Overview
This dataset is based on the MSMARCO Passage dataset and includes all the queries which have a positive passage in the original dataset (there are additional queries with no positive passages which we do not use). Each query has the known positive passage as well as 200 additional passages. These additional passages may be unlabeled positives or negatives. Each associated passage has a score from cross-encoder/ms-marco-MiniLM-L-12-v2 which were used for distillation training. The passages are ordered by the teacher score.
Dataset Construction
The 200 associated passages for each query were found by retrieving the top 800 passages using an early version of Hypencoder and then (1) selecting the top 100 passages (2) another 100 from the bottom 700 passages uniformly at random. Additionally, the known relevant passage provided by MSMARCO is added.
All query-passage pairs were then scored with cross-encoder/ms-marco-MiniLM-L-12-v2 to get teacher scores.
Dataset Structure
{
"query": {
"id": query ID which is the same as the one used in MSMARCO,
"content": query text,
"tokenized_content": token IDs for bert-base-uncased,
},
"items": [
{
"id": passage ID which is the same as the one used in MSMARCO,
"content": passage text,
"tokened_content": token IDs for bert-base-uncased,
"score": Teacher score from 'cross-encoder/ms-marco-MiniLM-L-12-v2/',
"type": Has two values [null, "given"] if the passage was the provided positive by MSMARCO the type is "given" otherwise it is null.
},
]
}
Cite
If you use this dataset please cite our paper:
@misc{killingback2025hypencoderhypernetworksinformationretrieval,
title={Hypencoder: Hypernetworks for Information Retrieval},
author={Julian Killingback and Hansi Zeng and Hamed Zamani},
year={2025},
eprint={2502.05364},
archivePrefix={arXiv},
primaryClass={cs.IR},
url={https://arxiv.org/abs/2502.05364},
}
and the original MSMARCO paper:
@inproceedings{MSMARCO,
author = {Tri Nguyen and
Mir Rosenberg and
Xia Song and
Jianfeng Gao and
Saurabh Tiwary and
Rangan Majumder and
Li Deng},
editor = {Tarek Richard Besold and
Antoine Bordes and
Artur S. d'Avila Garcez and
Greg Wayne},
title = {{MS} {MARCO:} {A} Human Generated MAchine Reading COmprehension Dataset},
booktitle = {Proceedings of the Workshop on Cognitive Computation: Integrating
neural and symbolic approaches 2016 co-located with the 30th Annual
Conference on Neural Information Processing Systems {(NIPS} 2016),
Barcelona, Spain, December 9, 2016},
series = {{CEUR} Workshop Proceedings},
volume = {1773},
publisher = {CEUR-WS.org},
year = {2016},
url = {https://ceur-ws.org/Vol-1773/CoCoNIPS\_2016\_paper9.pdf},
timestamp = {Thu, 11 Apr 2024 13:33:56 +0200},
biburl = {https://dblp.org/rec/conf/nips/NguyenRSGTMD16.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
- Downloads last month
- 19
Size of the auto-converted Parquet files (First 5GB):
1.74 GB
Number of rows (First 5GB):
24,098
Estimated number of rows:
335,553