DBRX
DBRX.
Databricks Released 27 Mar 2024 15
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Databricks Mar 2024 DBRX
Open weights
No Epoch Capabilities Index score yet
- GPQA Diamond 32.9% #194 /219
- Vectara Hallucination Leaderboard (HHEM) 8.30% #27 /148
- MATH Level 5 11.7% #85 /94
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 | |
|---|---|---|---|---|
| Vectara Hallucination Leaderboard (HHEM) | 8.30% | #27 of 148 | Official leaderboard | |
| EQ-Bench v2 | 76.8 | #33 of 85 | Official leaderboard | |
| AlpacaEval 2.0 | 25.4% | #20 of 40 | Official leaderboard | |
| WildBench | 35.5 | #29 of 47 | Official leaderboard | |
| LMArena (Elo) | 1,196 | #285 of 333 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GPQA Diamond | 32.9% | #194 of 219 | Epoch AI | |
| MATH Level 5 | 11.7% | #85 of 94 | Epoch AI |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HumanEval+ | 70.1% | #19 of 46 | Official leaderboard | |
| MBPP+ | 55.8% | #23 of 35 | Official leaderboard |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Berkeley Function Calling Leaderboard (BFCL) | 39.2% | #61 of 67 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Anthropic Red Team | 76.6% | #74 of 75 | Official leaderboard | |
| HELM Safety | 62.8 | #74 of 75 | Official leaderboard | |
| SimpleSafetyTests | 53.5% | #74 of 75 | Official leaderboard | |
| AIR-Bench 2024 | 25.4% | #82 of 82 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HELM Lite | 28.9 | #61 of 71 | Official leaderboard |