# The Multivac — Evaluation Report

**Evaluation ID:** EVAL-20260207-140934
**Date:** Feb 03, 2026
**Category:** code
**Question ID:** CODE-004

---

## Question

This function works but is O(n³). Optimize it to O(n log n) or better while maintaining correctness.

```python
def find_triplets_with_sum(arr, target_sum):
    """Find all unique triplets in arr that sum to target_sum"""
    n = len(arr)
    result = []
    
    for i in range(n):
        for j in range(i + 1, n):
            for k in range(j + 1, n):
                if arr[i] + arr[j] + arr[k] == target_sum:
                    triplet = sorted([arr[i], arr[j], arr[k]])
                    if triplet not in result:
                        result.append(triplet)
    
    return result
```

Explain your optimization approach and prove the new time complexity.

---

## Winner

**Claude Opus 4.5** (Anthropic)
- Winner Score: 9.64
- Matrix Average: 8.57
- Total Judgments: 90

---

## Rankings

| Rank | Model | Provider | Avg Score | Judgments |
|------|-------|----------|-----------|----------|
| 1 | Claude Opus 4.5 | Anthropic | 9.64 | 8 |
| 2 | Claude Sonnet 4.5 | Anthropic | 9.50 | 7 |
| 3 | DeepSeek V3.2 | DeepSeek | 9.49 | 8 |
| 4 | Grok 3 (Direct) | xAI | 9.49 | 8 |
| 5 | Gemini 3 Flash Preview | Google | 9.31 | 8 |
| 6 | GPT-5.2-Codex | OpenAI | 9.23 | 8 |
| 7 | Grok Code Fast | xAI | 9.09 | 8 |
| 8 | Gemini 3 Pro Preview | Google | 6.92 | 8 |
| 9 | GLM-4-7 | Zhipu | 6.80 | 4 |
| 10 | MiniMax M2 | MiniMax | 6.28 | 3 |

---

## 10×10 Judgment Matrix

Rows = Judge, Columns = Respondent. Self-judgments excluded (—).

| Judge ↓ / Resp → | Claude Sonnet | Grok Code Fast | Claude Opus | Gemini 3 | Gemini 3 | MiniMax M2 | GLM-4-7 | DeepSeek V3.2 | GPT-5.2-Codex | Grok 3 |
|---|---|---|---|---|---|---|---|---|---|---|
| Claude Sonnet | — | 8.8 | 9.8 | 7.0 | 9.6 | 0.0 | 0.0 | 9.8 | 8.3 | 9.2 |
| Grok Code Fast | 9.6 | — | 9.8 | 8.3 | 9.8 | 2.0 | 2.0 | 9.8 | 9.8 | 9.8 |
| Claude Opus | 9.3 | 9.3 | — | 7.7 | 9.6 | 0.0 | 0.0 | 9.6 | 9.4 | 9.3 |
| Gemini 3 | 0.0 | 0.0 | 0.0 | — | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| Gemini 3 | 9.8 | 9.8 | 10.0 | 8.4 | — | 0.0 | 0.0 | 9.8 | 9.8 | 9.8 |
| MiniMax M2 | 9.8 | 9.8 | 9.8 | 5.5 | 8.6 | — | 8.6 | 9.6 | 9.1 | 9.8 |
| GLM-4-7 | 0.0 | 9.6 | 9.8 | 0.0 | 9.6 | 0.0 | — | 9.8 | 9.8 | 9.6 |
| DeepSeek V3.2 | 9.6 | 9.0 | 10.0 | 7.9 | 9.6 | 9.8 | 9.6 | — | 9.1 | 9.6 |
| GPT-5.2-Codex | 8.8 | 7.8 | 8.8 | 3.6 | 8.8 | 0.0 | 0.0 | 8.8 | — | 8.8 |
| Grok 3 | 9.6 | 8.6 | 9.2 | 7.0 | 9.0 | 7.0 | 7.0 | 8.8 | 8.6 | — |

---

## Methodology

- **10×10 Blind Peer Matrix:** All models answer the same question, then all models judge all responses.
- **5 Criteria:** Correctness, completeness, clarity, depth, usefulness (each scored 1–10).
- **Self-judgments excluded:** Models do not judge their own responses.
- **Weighted Score:** Composite of all 5 criteria.

---

## Citation

The Multivac (2026). Blind Peer Evaluation: CODE-004. app.themultivac.com

## License

Open data. Free to use, share, and build upon. Please cite The Multivac when using this data.

Download raw JSON: https://app.themultivac.com/api/evaluations/EVAL-20260207-140934/results
Full dataset: https://app.themultivac.com/dashboard/export