Benchmark

MRCR 1M (pointwise) Benchmark

What is MRCR 1M (pointwise)?

MRCR 1M (pointwise) is a variant of the Multi-Round Coreference Resolution benchmark that uses pointwise evaluation for ultra-long contexts (~1M tokens). This…

Modality
text
Evaluates
long context, reasoning, general
Openness
unknown
Importer
LLM Stats
Reported scores
1

Source provenance

MRCR 1M (pointwise) results and reported scores

LLM Stats

ModelOrganizationReported valueReportedEvidence
Gemini 2.5 ProGoogle0.8292025-05-20evidence

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.

Which sources cite MRCR 1M (pointwise)?

Related long context, reasoning, general benchmarks