reasoning benchmark
CLUEWSC
CLUEWSC2020 is the Chinese version of the Winograd Schema Challenge, part of the CLUE benchmark. It focuses on pronoun disambiguation and coreference resolution, requiring models to determine which noun a pronoun refers to in a sentence. The dataset contains 1,244 training samples and 304 development samples extracted from contemporary Chinese literature.
Updated Aug 7, 2026
Published models3
Registry coverage3
MetricScore
EvidenceB
CLUEWSC leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MA | 91.4% | 100.0% | 3 | C | |
| 2 | DE | 90.9% | 50.0% | 3 | C | |
| 3 | BA | 48.6% | 0.0% | 3 | C |
Score distribution
Top published rows on the benchmark's original scale.
CLUEWSC highlights
The top published results on this benchmark's own scale.
What is CLUEWSC?
Definition and scoring fields from the benchmark registry.
CLUEWSC2020 is the Chinese version of the Winograd Schema Challenge, part of the CLUE benchmark. It focuses on pronoun disambiguation and coreference resolution, requiring models to determine which noun a pronoun refers to in a sentence. The dataset contains 1,244 training samples and 304 development samples extracted from contemporary Chinese literature.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- CLUEWSC
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- cluewsc|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about CLUEWSC.
Which model scores highest on CLUEWSC?
Kimi-k1.5 is currently ranked first with 91.4%.
What does CLUEWSC measure?
CLUEWSC2020 is the Chinese version of the Winograd Schema Challenge, part of the CLUE benchmark. It focuses on pronoun disambiguation and coreference resolution, requiring models to determine which noun a pronoun refers to in a sentence. The dataset contains 1,244 training samples and 304 development samples extracted from contemporary Chinese literature.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
3 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.