Benchmark
MRCR 128K (2-needle)
MRCR (Multi-Round Coreference Resolution) at 128K context length with 2 needles. Models must navigate long conversations to reproduce specific model…
- Modality
- text
- Categories
- long_context, reasoning, general
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 1
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| MiniCPM-SALA | OpenBMB | 0.2862 | 2026-02-11 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.