LiT (Lost in Translation) Robustness - Syntactic Complexity and Garden Paths: leaderboard
Metric: MQM>=80 rate (%): share of round-trip translations whose back-translation to English scores at least 80 on an MQM error scale (minor -1, major -5, critical -25) from a Grok 4.1 Fast judge comparing it with the English source; each passage is translated serially through a four-language sequence and back, averaged over eight sequences (high, medium and low-resource languages); robustness subset: garden-path and complex syntax; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GLM-5 | 63.5 |
| 2 | Qwen 3.5 397B A17B | 62.5 |
| 3 | Qwen 3 235B A22B 2507 (Thinking) | 51 |
| 4 | Kimi K2 (Thinking) | 51 |
| 5 | GLM-4.7 | 46.9 |
| 6 | Gemma 4 31B (IT) (Thinking) | 46.9 |
| 7 | Gemma 4 31B (IT) | 44.8 |
| 8 | DeepSeek V3.2 Exp (Non-reasoning) | 44.8 |
| 9 | GLM-5 (Non-reasoning) | 43.8 |
| 10 | MiniMax-M2.5 | 42.7 |
| 11 | Qwen 3.5 397B A17B (Non-reasoning) | 41.7 |
| 12 | DeepSeek V3.2 Exp (Thinking) | 40.6 |
| 13 | Qwen 3.5 35B A3B | 37.5 |
| 14 | Kimi K2 | 31.5 |
| 15 | MiMo-V2-Flash (Non-reasoning) | 30.2 |
Interactive version: theaggregate.ai/benchmark?slug=lit-lost-in-translation-robustness-syntactic-complexity-and-garden-paths · How It Works · Data refreshed daily, snapshot 2026-10-07.