You Don't Need to Run Every Eval
cs.LG
Submitted: 2026-06-22
Updated: 2026-08-26
Comments: 49 pages, 14 figures, 18 tables
Code: https://github.com/microsoft/benchpress
Project page: https://microsoft.github.io/benchpress
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Scales++: Compute Efficient Evaluation Subset Selection with Cognitive Scales Embeddings
- Revealing the structure of language model capabilities
- Anticipatory Evaluation of Language Models
- Do These LLM Benchmarks Agree? Fixing Benchmark Evaluation with BenchBench
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks