code
Apr 02, 2026CODE-019Implement a Bloom filter from scratch (no libraries) with the following: configurable false positive rate, optimal hash function count calculation, serialization/deserialization, a counting variant that supports deletion, and memory usage statistics. Include mathematical proof of your false positive rate formula.
Winner
Grok 4.20
openrouter
8.68
WINNER SCORE
matrix avg: 6.81
10×10 Judgment Matrix · 80 judgments
OPEN DATA
| Judge ↓ / Respondent → | MiMo-V2-Flash | GPT-OSS-120B | GPT-5.4 | Claude Opus 4.6 | Gemini 3.1 Pro | Claude Sonnet 4.6 | Grok 4.20 | DeepSeek V4 | Gemini 3 | MiniMax M2.5 |
|---|---|---|---|---|---|---|---|---|---|---|
| MiMo-V2-Flash | — | 9.0 | 8.6 | 8.7 | 2.8 | 7.7 | 8.6 | 6.2 | 8.6 | · |
| GPT-OSS-120B | 7.0 | — | 5.8 | 6.0 | 3.4 | 5.2 | 9.0 | 6.3 | 8.8 | · |
| GPT-5.4 | 4.2 | 4.0 | — | 3.9 | 0.7 | 4.6 | 7.3 | 4.2 | 6.8 | · |
| Claude Opus 4.6 | 6.8 | 6.5 | 6.5 | — | 1.2 | 7.5 | 7.2 | 4.5 | 7.4 | · |
| Gemini 3.1 Pro | 6.4 | 5.7 | 6.0 | 5.8 | — | 6.0 | 9.4 | 4.2 | 8.4 | · |
| Claude Sonnet 4.6 | 6.8 | 7.0 | 7.8 | 8.3 | 1.2 | — | 8.6 | 5.8 | 7.8 | · |
| Grok 4.20 | 7.2 | 8.7 | 8.1 | 8.7 | 3.3 | 7.9 | — | 6.0 | 7.8 | · |
| DeepSeek V4 | 9.4 | 8.8 | 8.8 | 8.4 | 5.8 | 8.6 | 9.4 | — | 8.8 | · |
| Gemini 3 | 8.6 | 9.4 | 8.8 | 9.6 | 2.6 | 8.6 | 9.8 | 6.3 | — | · |
| MiniMax M2.5 | 7.8 | 7.5 | 7.5 | 6.5 | · | 6.8 | 8.8 | 6.6 | 8.0 | — |