Gaming Without an Attacker: Benchmark Fingerprinting in LLM-Driven Search Under Selection Pressure
Víctor Gallego
cs.LG, cs.AI
Submitted: 2026-08-09
Updated: 2026-08-11
Comments: Published as a conference paper at AI Measurement Science Workshop @ COLM 2026
Code: https://github.com/vicgalle/kernel-fingerprinting
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Training on the Test Task Confounds Evaluation and Emergence
- Metal-Sci: A Scientific Compute Benchmark for Evolutionary LLM Kernel Search on Apple Silicon
- Discovering Cooperative Pipelines: Autoresearch for Sequential Social Dilemmas
- Categorizing Variants of Goodhart's Law
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
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: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks