Related papers: Mutation Strength Adaptation of the $(\mu/\mu_I, \…
The capacity of meta-learning algorithms to quickly adapt to a variety of tasks, including ones they did not experience during meta-training, has been a key factor in the recent success of these methods on few-shot learning problems. This…
The fitness landscape - the mapping between genotypes and fitness - determines properties of the process of adaptation. Several small genetic fitness landscapes have recently been built by selecting a handful of beneficial mutations and…
We consider a stochastic individual-based model of adaptive dynamics for an asexually reproducing population with mutation, with linear birth and death rates, as well as a density-dependent competition. To depict repeating changes of the…
It is well-known that an extremely accurate parametrization of the growth function of matter density perturbations in $\Lambda$CDM cosmology, with errors below $0.25 \%$, is given by $f(a)=\Omega_{m}^{\gamma} \,(a)$ with $\gamma \simeq…
Reconstructing the evolution of post-common-envelope binaries (PCEBs) can constrain current prescriptions of common-envelope (CE) evolution. Analyzing a new sample of PCEBs we derive constraints on one of the most important parameters in…
Experimental robot optimization often requires evaluating each candidate policy for seconds to minutes. The chosen evaluation time influences optimization because of a speed-accuracy tradeoff: shorter evaluations enable faster iteration,…
The integration of Large Language Models (LLMs) with evolutionary computation (EC) has introduced a promising paradigm for automating the design of metaheuristic algorithms. However, existing frameworks, such as the Large Language Model…
Despite its popularity, several recent works question the effectiveness of MAML when test tasks are different from training tasks, thus suggesting various task-conditioned methodology to improve the initialization. Instead of searching for…
Self-adaptive parameters are increasingly used in the field of Evolutionary Robotics, as they allow key evolutionary rates to vary autonomously in a context-sensitive manner throughout the optimisation process. A significant limitation to…
Machine learning applications on signals such as computer vision or biomedical data often face significant challenges due to the variability that exists across hardware devices or session recordings. This variability poses a Domain…
The common-envelope (CE) phase is an important stage in the evolution of binary stellar populations. The most common way to compute the change in orbital period during a CE is to relate the binding energy of the envelope of the Roche-lobe…
It is known (see e.g. Weibull (1995)) that ESS is not robust against multiple mutations. In this article, we introduce robustness against multiple mutations and study some equivalent formulations and consequences.
Continual Test-Time Adaptation (CTTA) generalizes conventional Test-Time Adaptation (TTA) by assuming that the target domain is dynamic over time rather than stationary. In this paper, we explore Multi-Modal Continual Test-Time Adaptation…
We compile a holistic scenario for molecular cloud (MC) evolution and control of the star formation efficiency (SFE), and present a first set of numerical tests of it. A {\it lossy} compressible cascade can generate density fluctuations and…
The quadratic complexity of the attention mechanism in Transformer models has motivated the development of alternative architectures with sub-quadratic scaling, such as state-space models. Among these, Mamba has emerged as a leading…
In this paper, we analyze the behavior of the multivariate symmetric uncertainty (MSU) measure through the use of statistical simulation techniques under various mixes of informative and non-informative randomly generated features.…
The early planetesimal growth proceeds through a sequence of sticking collisions of dust agglomerates. Very uncertain is still the relative velocity regime in which growth rather than destruction can take place. The outcome of a collision…
We present new fitting functions for the velocity divergence auto- and cross-power spectra, $P_{\theta\theta}(k)$ and $P_{\delta\theta}(k)$, calibrated on gravity-only $N$-body simulations. By applying the Evolution Mapping framework, we…
Modern machine learning uses more and more advanced optimization techniques to find optimal hyper parameters. Whenever the objective function is non-convex, non continuous and with potentially multiple local minima, standard gradient…
This paper focuses on the restart strategy of CMA-ES on multi-modal functions. A first alternative strategy proceeds by decreasing the initial step-size of the mutation while doubling the population size at each restart. A second strategy…