Taplio - Build a growth routine on LinkedIn
Mini toolsARC Challenge
AI2 Reasoning Challenge (Challenge set)
Grade-school science multiple-choice, hard subset. Saturated by frontier models but still in many evaluation harnesses.
Knowledge
Text
accuracy
Max 100.0%
Released Mar 2018
Saturated
Possibly contaminated
10
Results
9
Models scored
554.0%
Top: Gemma 2
91.3%
Median
Best results
Top primary scores; one row per model.
Frontier over time
Each dot is one model result; the line traces the running best score.
All results
Showing one canonical row per model. Show all configurations
| # | Model | Score | Conditions | Eval date | Source | Flags |
|---|---|---|---|---|---|---|
| 1 | Gemma 2 | 554.0% | — | Feb 25, 2025 | self reported | primary |
| 2 | Claude Opus 3 | 96.4% | 25-shot | Oct 22, 2024 | self reported | primary |
| 3 | Nemotron 3 Super | 96.1% | 25-shot | Apr 3, 2026 | self reported | primary |
| 4 | Nova Pro | 94.8% | 0-shot | Dec 3, 2024 | self reported | primary |
| 5 | Nova Lite | 92.4% | 0-shot | Dec 3, 2024 | self reported | primary |
| 6 | Nova Micro | 90.2% | 0-shot | Dec 3, 2024 | self reported | primary |
| 7 | Llama 3.2 | 78.6% | 0-shot | Oct 22, 2024 | self reported | primary |
| 8 | Mixtral 8x7B | 59.7% | — | Dec 1, 2023 | self reported | primary |
| 9 | Mixtral 8x7B | 59.7% | — | Jan 8, 2024 | self reported | primary |
| 10 | Mistral 7B | 55.6% | — | Sep 1, 2023 | self reported | primary |
