Research map / Large language models / Retrieval and knowledge
Mapping Overlaps in Benchmarks through Perplexity in the Wild
Siyang Wu, Honglin Bao, Sida Li, Ari Holtzman, James Evans
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use token-level instead of document-level perplexity (ablation) | signature extraction · GSM8K | adjusted R2: 0.826 -> 0.927 | Improved |
| use token-level instead of chunk-level perplexity (ablation) | signature extraction · GSM8K | adjusted R2: 0.903 -> 0.927 | Improved |
| use token-level instead of chunk-level perplexity (ablation) | signature extraction · MBPP | adjusted R2: 0.849 -> 0.885 | Improved |
| use token-level instead of chunk-level perplexity (ablation) | signature extraction · MMLU | adjusted R2: 0.551 -> 0.454 | Hurt |
| use token-level instead of document-level perplexity (ablation) | signature extraction · MMLU | adjusted R2: 0.185 -> 0.454 | Improved |
| use token-level instead of chunk-level perplexity (ablation) | signature extraction · TruthfulQA | adjusted R2: 0.414 -> 0.505 | Improved |
| use signature-level instead of performance-level overlap | benchmark overlap analysis · within vs cross family/format overlap | overlap difference: performance overlap about 0.8 vs signature near 0 | Improved |
Closest papers by meaning
- metabench - A Sparse Benchmark of Reasoning and Knowledge in Large Language Models (ICLR 2025 · similarity 0.91)
- BenchmarkCards: Standardized Documentation for Large Language Model Benchmarks (NeurIPS 2025 · similarity 0.90)
- Potemkin Understanding in Large Language Models (ICML 2025 · similarity 0.89)
- What is Wrong with Perplexity for Long-context Language Modeling? (ICLR 2025 · similarity 0.89)
- Exploring the Limits of Sub-Billion Language Model Reasoners with Open Training Recipes (ICLR 2026 · similarity 0.88)
- Measuring what Matters: Construct Validity in Large Language Model Benchmarks (NeurIPS 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
