When Can Old Evaluations Certify a New Model? Label-Efficient Release Decisions under Evaluator Drift
cs.LG
Submitted: 2026-09-26
Updated: 2026-09-26
Terminology
Sources
- PPI++: Efficient Prediction-Powered Inference
- Best Arm Identification with LLM Judges and Limited Human
- Pay Only for Disagreement: Certified No-Regression Verdicts for Model Updates with Matching Label-Complexity Bounds
- Who Drifted: the System or the Judge? Anytime-Valid Attribution in LLM Evaluation Pipelines
- When the Judge Changes, So Does the Measurement: Auditing LLM-as-Judge Reliability
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