Yi-34B
Yi-34B ranks first on SuperCLUE-Safety.
23
Benchmarks with a score
1
Ranked first, out of five or more models
1
In the top three
01.AI Nov 2023 Yi-34B
Open weights
No Epoch Capabilities Index score yet
- GPQA Diamond 14.7% #227 /229
- OlympicArena 18% #19 /25
- MathBench 31.6% #17 /21
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MMLU | 76.3% | #17 of 85 | Claim in Yi: Open Foundation Models by 01.AI, via Epoch AI | |
| BIG-Bench Hard | 71.7% | #10 of 45 | Claim in Yi: Open Foundation Models by 01.AI, via Epoch AI | |
| AlpacaEval | 94.1% | #5 of 19 | Official leaderboard | |
| AlpacaEval 2.0 | 27.2% | #19 of 47 | Official leaderboard | |
| MuSR | 41.2% | #50 of 98 | Official leaderboard | |
| EQ-Bench v2 | 71.6 | #50 of 85 | Official leaderboard | |
| MMLU-Pro | 43% | #154 of 212 | Official leaderboard | |
| LMArena (Elo) | 1,184 | #291 of 335 | Official leaderboard | |
| BABILong | 39.6% | #6 of 6 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GSM8K | 76% | #20 of 70 | Claim in Yi: Open Foundation Models by 01.AI, via Epoch AI | |
| OlympicArena | 18% | #19 of 25 | Official leaderboard | |
| MathBench | 31.6% | #17 of 21 | Official leaderboard | |
| MATH Level 5 | 5.15% | #92 of 94 | Epoch AI run | |
| GPQA Diamond | 14.7% | #227 of 229 | Epoch AI run |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GTA | 3.21% | #13 of 15 | Official leaderboard |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MedQA | 65.6% | #36 of 71 | Official leaderboard |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| TemporalBench | 18.5% | #7 of 10 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| SuperCLUE-Safety | 89.3 | #1 of 7 | Official leaderboard | |
| WMDP Biology | 75.3% | #1 of 2 | Official leaderboard | |
| AIR-Bench 2024 | 53.6% | #59 of 82 | Official leaderboard | |
| WMDP Cyber | 49.7% | #2 of 2 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HELM Lite | 57.0 | #30 of 71 | Official leaderboard | |
| Open LLM Leaderboard v2 | 22.4 | #50 of 99 | Official leaderboard |