.
La vertiente occidental de la Circumpuna de Atacama presenta importantes yacimientos de cobre y concentra la mayor cantidad de evidencias mineras prehispánicas de los Andes del Sur. No obstante, no ocurre lo mismo con las evidencias metalúrgicas, siendo estas poco conocidas. Las prácticas metalúrgicas de las comunidades pastoras-caravaneras y mineras de tierras altas de la región (2400-4500 m snm) tuvieron lugar durante los períodos Intermedio Tardío y Tardío (1000-1536 D.C.), extendiéndose desde Collahuasi por el norte hasta Catarpe por el sur. En este trabajo caracterizamos las operaciones metalúrgicas del área en cuanto a los minerales reducidos, el metal obtenido, los combustibles y las estructuras de combustión utilizadas, permitiéndonos proponer una tradición metalúrgica de la vertiente occidental de la Circumpuna de Atacama orientada a la producción de cobre metálico sin alear.
Multimodal continual learning with Mixture-of-Experts (MoE) typically relies on task identifiers to route inputs, limiting deployment when task boundaries are unknown. We propose ComplementarityMoE, which replaces task-ID routing with a Barlow Twins loss adapted for complementarity: instead of enforcing view-invariance ( τ =1.0 ), we target partial correlation ( τ<1.0 ) between router projections, driving experts toward orthogonal feature subspaces without requiring task labels. The architecture freezes a PerceiverIO encoder after the first task and trains only LoRA-based experts and the router on subsequent tasks, enabling positive backward transfer. We scale this approach from 2 to 10 sequential tasks on CMU-MOSEI and uncover a key finding: the optimal complementarity level reverses with task count. At 2 tasks, τ =0.5 is optimal (51.93 τ =0.2 dominates (46.31 τ =0.5 ), outperforming D-MoLE, CL-MoE, ProgLoRA, EWC, and naive fine-tuning with statistical significance ( p<0.05 , 10 seeds). Routing analysis reveals that the router converges to near-perfect expert specialization (>99
In this study, 12 selected carved lapis lazuli amulets from various periods of ancient Egyptian history, from the New Kingdom to the Graeco-Roman Period, were analysed non-invasively to determine the provenance of their raw material. The amulets are part of the Museo Egizio of Turin (Italy) collection and include five samples retrieved at the ancient site of Heliopolis (Egypt). To ascertain the lapis lazuli provenance, in-air ion beam analyses (IBA) were employed to measure trace element content and luminescent properties of the various mineral species within the rock (diopside, pyrite, and calcite). Results were compared with those from a large set of reference rocks (almost seventy samples) from five mining areas in present-day Chile, Afghanistan, Siberia, Tajikistan, and Myanmar. Multivariate statistical analysis, in particular principal component analysis, was largely employed to drive the provenance assessment. The results enabled clear provenance identification in most cases. It has been demonstrated that the lapis lazuli of 10 amulets originated from Afghan quarries, reinforcing the archaeological hypothesis of a continuous primary supply of Egypt from this region during the II-I millennium BCE. This result is particularly significant for the Heliopolis amulets, which are among the few Egyptian lapis lazuli artefacts from a known archaeological context analysed in literature for provenance studies. The remaining two amulets show more complex but potentially interesting results, especially an amulet from the Graeco-Roman Period that seems to be a rare example of carved object from the Mediterranean area whose raw material does not show chemical–physical compatibility with the Afghan source. This work represents one of the few provenance studies to include such a significant set of lapis lazuli artefacts associated with the material culture of ancient Egypt, spanning different historical periods, and systematically compared with a large database of geological reference rocks.
X-ray fluorescence (XRF) spectroscopy is a key modality for material analysis in cultural heritage. However, automated learning from XRF spectra remains challenging: XRF spectra are complex one-dimensional signals composed of sharp elemental peaks, broader structures, and background variations that are not taken into account by existing learning-based models. This paper introduces XRFormer, a transformer architecture tailored to XRF spectra through a multiscale convolutional tokenizer that injects locality and multi-resolution inductive biases before global self-attention. The tokenizer progressively reduces spectral resolution while increasing embedding dimensionality, and the resulting token sequence is processed by a standard transformer encoder. We further investigate self-supervised pretraining for XRF representation learning using Masked Spectral modeling (MSM) and a physics-informed Peak Presence Prediction (PPP) objective. Experiments on the Pigments Checker STANDARD v.5 dataset for pigment identification and unmixing show that XRFormer consistently outperforms ViT, SpectralFormer (with and without CAF), and a 1D-CNN baseline for pigment identification. For pigment unmixing, XRFormer achieves robust abundance estimation while maintaining significantly higher parameter efficiency than SpectralFormer, operating at a lower token resolution (128 vs. 512 tokens) and with less than half the number of parameters (1.5M vs. 3.37M). MSM yields consistent gains across both tasks, while PPP further enhances performance for both identification and unmixing when tuned with an appropriate peak prominence. These results highlight multiscale, modality-aware tokenization as an effective and parameter efficient foundation for transformer-based XRF modeling under data-limited conditions. A GitHub repository is provided at https://github.com/sofiane1010/XRFormer.
Se presenta una síntesis de información geoarqueológica compilada por diferentes autores para los distritos mineros del Desierto de Atacama, entre ~20°0’00’’S y ~24°0’00’’S, incluyendo los distritos mineros de Collahuasi, El Abra, Chuquicamata, Cerro Verde y San Bartolo. Algunos de estos distritos presentan evidencia directa de explotación prehispánica o bien se encuentran próximos a sitios arqueológicos relevantes para la problemática minero-metalúrgica prehispánica. Se integra información mineralógica y geoquímica inédita de muestras específicas, obtenida a partir de análisis realizados mediante 10 técnicas analíticas distintas y complementarias, con el objetivo de construir una cartografía mineral del área. Esta cartografía permite reconstruir la cronología del uso del cobre en sitios arqueológicos del norte de Chile, desde el Periodo Arcaico Tardío hasta el Periodo Tardío.
