Dataset · Hugging Face

uwm-software-engineering/complexity-kink-research

Name: complexity-kink-research
Creator: uwm-software-engineering
License: cc-by-4.0
Keywords: task_categories:text-generation, annotations_creators:machine-generated, annotations_creators:expert-generated, source_datasets:nvidia/OpenCodeInstruct, language:en, license:cc-by-4.0, size_categories:100K<n<1M, format:parquet, modality:tabular, modality:text, library:datasets, library:pandas, library:polars, library:mlcroissant, arxiv:2504.04030

Complexity Kink: prompt-side complexity, LLM code generations, and test outcomes This dataset accompanies the NeurIPS 2026 Evaluations & Datasets track paper The Complexity Kink: LLM Rubric Instruments for Causal Inference on Code Generation Reliability (Michael Hernandez and Tian Zhao, University of Wisconsin-Milwaukee). Code and analysis: https://github.com/uwm-se/ComplexityKink. It scores the structural complexity of 5,000 Python programming tasks from the prompt alone, never… See the full description on the dataset page: https://huggingface.co/datasets/uwm-software-engineering/complexity-k

View on Hugging FaceReport an issue

Author / organization	uwm-software-engineering
License (as declared)	cc-by-4.0
Downloads	342 (as reported by Hugging Face)
Likes	1
Created	16 October 2026
Last updated at source	17 October 2026

codecode-generationpythonllm-evaluationllm-as-a-judgebenchmarkcyclomatic-complexity

Embed this entry

<iframe src="https://silytics.si/embed/item/00f2b7ad-65c5-49b3-8762-8f5f2700a7d8" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 2 days

342 now (+342 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Connected in the knowledge graph

kwamemarfo916/KallyModel · used to train

Explore this entry in the graph

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 20 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.