arXiv:2607. 01539v1 Announce Type: cross Abstract: This paper studies whether the Metabolic Multi-Agent Optimizer (MMAO) can act as a credible outer-loop optimizer for classification model selection.
arXiv:2606. 01566v1 Announce Type: new Abstract: Small-to-medium scientific datasets place machine learning pipelines under two compounding pressures.
The paper introduces the Min-Marginal Power Index (MinPI), a game-theoretic feature selection method that labels features as harmful, neutral, or decisive based on whether adding them to a coalition drops utility below a threshold. MinPI enables conservative feature removal without a score cutoff, and its application to six benchmark datasets removed 12.5% of features with negligible loss in held‑out utility. The method also improves multi‑objective search performance by prefiltering features, especially in datasets with many removable features.
The paper investigates two strategies for improving large language model (LLM) evaluation: specialized judge weights and rule‑based deferral policies. Experiments on nearly 100,000 rubric‑conditioned samples show that correct rubrics boost accuracy, while incorrect ones hurt it, and that splitting training data into criterion‑specific experts can severely degrade performance unless the experts are warm‑started from a unified model. The authors demonstrate that lightweight deferral cascades can match or exceed the accuracy of larger standalone judges at a fraction of the compute cost, and they provide practical design rules for building efficient, reliable LLM evaluators.
arXiv:2606. 21641v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have been proposed as hyperparameter-optimization (HPO) advisors that "warm-start" search from prior knowledge, proposing strong configurations in very few evaluations.
SelectArena is a new benchmark for feature selection on real-world tabular data, evaluating 15 popular methods across 71 curated datasets for both classification and regression tasks. The study uses standardized protocols and modern foundation models, revealing trade-offs among predictive performance, runtime, stability, and validity. Tree-based methods (CART, RFImportance) and the F-Test emerge as Pareto-optimal choices, while search-based methods perform poorly in speed and stability.
The paper introduces TGL-NSGA-II, a low‑fidelity framework that uses a pretrained teacher to stratify samples by difficulty and class, then applies a short knowledge‑distillation step (KD‑Lite) before scoring candidates on a stratified evaluation set. The teacher‑guided scores are fused with a Gaussian‑process surrogate to select candidates for full evaluation, and the method is evaluated on keyword spotting and bird‑call classification tasks. Results show high Kendall‑τ values (0.74 and 0.62), a 41% reduction in proxy‑score variance, and improved hypervolume and false‑positive rates compared to full NSGA‑II, while running 2.2× faster under a constrained evaluation budget.
ESPO (Error-Structured Prompt Optimization) addresses prompt bloat in evolutionary prompt optimizers by separating optimization into Diagnose, Propose, and Select phases. It clusters training errors, generates diverse candidates, and applies bootstrap stability selection, achieving a 3.76‑point accuracy gain over GEPA on seven NLP benchmarks while producing 47% shorter prompts. Cross‑model tests on four additional student models confirm ESPO’s superior average accuracy, notably improving Qwen3 GSM8K from 15.00% to 91.40%.
The paper introduces WHALE, a method that alternates between updating a language model’s weights and searching for a better harness (the code that manages context and control flow). By iteratively fine‑tuning the model under the current harness and then optimizing the harness under the updated model, WHALE improves performance across search QA, math reasoning, and chess puzzles, outperforming weight‑only, harness‑only, and Fast‑Slow Training by 4.15–24.38 percentage points in mean@8 accuracy. The approach uses either fixed phase lengths or an adaptive patience rule to decide when to switch phases, and the authors provide code on GitHub.
arXiv:2608. 02305v1 Announce Type: new Abstract: Active feature acquisition (AFA) asks which unobserved feature to measure next for each test instance under a budget.
TIGER is a time‑series classification method that uses a small set of three generic classifiers applied to four different representation families, producing twelve base learners. The predictions are stacked into a meta‑feature matrix and an adaptive meta‑classifier—choosing between a weighted hard majority vote and the pretrained TabICLv2 model—selects the best rule per dataset based on training sample size. On a 142‑dataset UCR benchmark, TIGER achieves the highest mean accuracy, balanced accuracy, and F1‑score among six compared algorithms, outperforming each constituent method and demonstrating strong generalization with a single hyperparameter.
Agentic systems have widened the gap between producing candidate outputs and reviewing them. This paper asks a practical architectural question: should domain specialization be built into an evaluator's weights, or into the rule that decides when its judgment can be trusted?
arXiv:2606. 00862v1 Announce Type: cross Abstract: Surrogate-assisted evolutionary algorithms (SAEAs) have been widely used for expensive black-box optimization problems.