category stringclasses 8
values | key int64 4 99 | question stringclasses 10
values | ground_truths stringclasses 10
values | misc stringclasses 5
values |
|---|---|---|---|---|
scifacts-geo | 55 | Find the paper(s) that use all of the datasets mentioned below:
American Community Survey, Flooding Data from FEMA: Harvey Flood Depths Grid, Food Access Research Atlas by USDA, Streetlight Data, POI data by SafeGraph
Instructions:
1. You must respond with a list JSON object(s) in this format:
{
"paper_title":... | [[{"paper_title": ["Characterizing Equitable Access to Grocery Stores During Disasters Using Location-based Data"]}]] | {"sources": [], "eval_info": {}} |
scifacts-materials | 76 | Find the material(s) that satisfy every one of the listed measured properties.
Instructions
1. In your response, provide a list of JSON object(s) in the following format:
[
{
"material": "<material name>",
"inference_basis": "<brief explanation of how the listed properties match this material>",
... | [[{"paper_title": ["Hydrothermally Grown ZnO Micro/Nanotube Arrays and Their Properties"], "material": ["zinc oxide"]}]] | {"sources": [], "eval_info": {}} |
entities | 10 | Provide a comprehensive list of all animated films (excluding anime) that meet the following criteria:
1. The film must fall under the category of animation but must not be classified as anime.
2. The film must have a Tomatometer rating of 95% or higher on Rotten Tomatoes.
3. The film’s runtime must be between 80 an... | [["Beauty and the Beast (1991)", "Shaun the Sheep Movie", "Snow White and the Seven Dwarfs", "A Boy Named Charlie Brown", "The Big Bad Fox and Other Tales...", "The King and the Mockingbird", "Batman: The Long Halloween, Part One", "Steven Universe: The Movie", "Predator: Killer of Killers", "Flow (2024)", "Marvel Risi... | {"sources": [], "eval_info": {}} |
entities | 4 | Provide a comprehensive list of all people that meet the following criteria:
1. He/She should have won atleast 3 IMO medals for Korea.
2. Atleast one of these medals should have been won in an year when Korea was ranked a minimum of 5(<=5) during IMO.
Present the final result as a JSON array of strings, where each s... | [["Youngbeom Jin", "Woojin Choi", "Gyudong Lee", "Sehun Kim", "Jimin Kim", "Dong Ryul Kim", "Whan Ghang", "Junhwi Bae", "Yuchan Jung"]] | {"sources": [], "eval_info": {}} |
novel-datasets-identification | 31 | Identify a publicly available dataset of long-form in-the-wild videos that are segmented into scenes and richly annotated with both content descriptions and user engagement signals. The dataset should include multimodal information and support analysis at the video level as well as the scene level.
Once you've found a... | [{"title": "Large Content and Behavior Models to Understand, Simulate, and Optimize Content and Behavior", "is_published": true, "venue": "ICLR", "year": 2024, "dataset_url": "https://huggingface.co/datasets/behavior-in-the-wild/content-behavior-corpus"}] | {"sources": ["https://arxiv.org/pdf/2309.00359"], "eval_info": {"main_claims": ["title"]}} |
novel-datasets-identi-extraction | 23 | I'm looking for a GPT-4 generated corpus of decisions rooted in quotidian life - think commuting, family squabbles or career decisions - each tagged against broad socio-psychological dimensions. I'm specifically interested in scenarios where the resolution hinges not on objective correctness, but instead on personal va... | [[{"framework": "Moral Foundations Theory", "most_biased_dimension": "Fairness"}, {"framework": "World Values Survey", "most_biased_dimension": "Self-Expression"}, {"framework": "Maslow's Hierarchy of Needs", "most_biased_dimension": "Self-Esteem"}, {"framework": "Plutchik's Wheel of Emotions", "most_biased_dimension":... | {"sources": ["https://openreview.net/pdf?id=PGhiPGBf47"], "eval_info": {"primary_keys": ["framework"], "main_claims": ["framework"]}} |
novel-datasets-peer | 38 | I want to compare existing 3D urban segmentation datasets based on real-world scenes that go beyond static 2D maps or flat segmentation overlays, and instead represent urban environments as richly annotated point clouds. How do they compare in terms of data acquisition methods—such as mobile laser scanning, terrestrial... | [[{"name": "Okland", "data_aquisition_method": "MLS", "area": "1.5 km", "scenes": 1, "points_million": 1.6}, {"name": "Semantic3D", "data_aquisition_method": "TLS", "area": "-", "scenes": 3, "points_million": 4000}, {"name": "Paris-Lille-3D", "data_aquisition_method": "MLS", "area": "1.94 km", "scenes": 2, "points_mill... | {"sources": ["https://arxiv.org/pdf/2305.02627"], "eval_info": {"primary_keys": ["name"], "ignore_keys": ["scenes"], "note": {"area": "tolerance: \u00b11 km or \u00b1 1 km\u00b2", "points_million": "tolerance: 10% of the ground truth value"}, "main_claims": ["name"]}} |
flights | 40 | In which flight incident did a commercial airliner perform an unusually high number of go-arounds before safely landing? For each attempt, add an entry using the following format:
```json
[
{
"time_utc": <timestamp>,
"attempt_number": <attempt_number>,
"airport": <airport_name>,
"ru... | [[{"time_(utc)": "2358", "attempt_#": "1", "airport": "Cochin", "runway_#": "27", "runway_aid_available": "ILS Cat I", "visibility_minima_(for_aid)": "650m RVR, 320ft DA", "actual_visibility": "3500m", "remaining_fuel_(estimate)": "4699 kg"}, {"time_(utc)": "0017", "attempt_#": "2", "airport": "Cochin", "runway_#": "27... | {"sources": ["https://asn.flightsafety.org/reports/2015/20150818_B738_VT-JFA.pdf"], "eval_info": {"primary_keys": ["time_utc", "attempt_number"], "note": {"time_utc": "\u00b15 minutes", "visibility_minima_for_aid": "\u00b1200m", "remaining_fuel_estimate": "\u00b1200 kg"}, "main_claims": ["time_utc", "attempt_number"]}} |
prior-art | 86 | _I have the following ideas for a research paper. Can you help identify if this has already been done or implied in full or in parts in other papers?
Give your answer as a **JSON with three fields: Paper title, link, and connection a field** that quotes exact sentences from the paper and parts of my ideas below to make... | [[{"title": "Exposing the Achilles' Heel: Evaluating LLMs Ability to Handle Mistakes in Mathematical Reasoning", "url": "https://openreview.net/forum?id=uDZ9d4UAUh"}, {"title": "Large Language Models Cannot Self-Correct Reasoning Yet", "url": "https://arxiv.org/abs/2310.01798"}, {"title": "SELF-[IN]CORRECT: LLMs Strugg... | {"sources": [], "eval_info": {}} |
prior-art | 99 | _I have the following ideas for a research paper. Can you help identify if this has already been done or implied in full or in parts in other papers?
Give your answer as a **JSON with three fields: Paper title, link, and connection a field** that quotes exact sentences from the paper and parts of my ideas below to make... | [[{"title": "Theoretical Analysis of Contrastive Learning in Vision-Language Model Pretraining: The Role of Synthetic Text Captions for Feature Alignment", "url": "https://openreview.net/forum?id=hgAAXdv8q8"}, {"title": "Understanding Multimodal Contrastive Learning and Incorporating Unpaired Data", "url": "https://arx... | {"sources": [], "eval_info": {}} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.