Ai2Arc
Dataset Card for "ai2_arc"
- Publisher
- Allen Institute for AI
- License
- cc-by-sa-4.0
- Published
- May 2026
Cite
Notes
Only stored in your browser.
Top score 0.0% by GPT-5 Nano - 2 models reporting (2 frontier)
Sample tasks
5from the eval dataset
An astronomer observes that a planet rotates faster after a meteorite impact. Which is the most likely effect of this increase in rotation?
- APlanetary density will decrease.
- BPlanetary years will become longer.
- CPlanetary days will become shorter.
- DPlanetary gravity will become stronger.
Show 4 more examples
A group of engineers wanted to know how different building designs would respond during an earthquake. They made several models of buildings and tested each for its ability to withstand earthquake conditions. Which will most likely result from testing different building designs?
- Abuildings will be built faster
- Bbuildings will be made safer
- Cbuilding designs will look nicer
- Dbuilding materials will be cheaper
The end result in the process of photosynthesis is the production of sugar and oxygen. Which step signals the beginning of photosynthesis?
- AChemical energy is absorbed through the roots.
- BLight energy is converted to chemical energy.
- CChlorophyll in the leaf captures light energy.
- DSunlight is converted into chlorophyll.
A physicist wants to determine the speed a car must reach to jump over a ramp. The physicist conducts three trials. In trials two and three, the speed of the car is increased by 20 miles per hour. What is the physicist investigating when he changes the speed?
- Athe control
- Bthe hypothesis statement
- Cthe dependent (responding) variable
- Dthe independent (manipulated) variable
An astronaut drops a 1.0 kg object and a 5.0 kg object on the Moon. Both objects fall a total distance of 2.0 m vertically. Which of the following best describes the objects after they have fallen a distance of 1.0 m?
- AThey have each lost kinetic energy.
- BThey have each gained the same amount of potential energy.
- CThey have each lost the same amount of potential energy.
- DThey have each gained one-half of their maximum kinetic energy.
Score history
2Top models
2Related tools
1Implementations, trainers, datasets and scaffolds linked to this eval.
FAQ
- What is Ai2Arc?
- Dataset Card for "ai2_arc"
- What is the current top score on Ai2Arc?
- The top reported score is 0.0% by GPT-5 Nano, across 2 models reporting (2 from frontier labs).
- How can a model improve its Ai2Arc score?
- Tools linked to Ai2Arc on Sophon include AI 2 ARC RL Env (Community) - RL environments, datasets, and scaffolds that target this eval.
- What license is Ai2Arc under?
- Ai2Arc is available under cc-by-sa-4.0.