WebDev Arena Preference 10K: LLM Battles for Web Development Tasks

10,000 real-world WebDev Arena battles involving 10 state-of-the-art large language models (LLMs). The dataset was created by lmarena-ai and was last updated on March 10, 2025. It is hosted on the Hugging Face platform.

Use Cases

• Training preference models for LLMs based on battle outcomes.
• Benchmarking LLM performance on web development-related prompts.
• Analyzing model preferences and rankings in a competitive evaluation setting.

Strengths

• Contains 10,000 real-world comparison instances.
• Involves 10 state-of-the-art LLMs for evaluation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

lmarena-ai via Hugging Face.

Collection Method

Likely contains pairwise comparison data (battles) between LLMs on web development tasks.

Freshness

Last updated 2025-03-10 19:36:45; freshness should be verified.