arXiv subjects

Daniel Gomes

Publications and source records attributed to Daniel Gomes.

5 recordsLinked to original sources

arXiv2026

Fine PT-PT Web: A High-Quality 41 Billion Tokens Data Collection of the European Portuguese Web

Curating Web corpora for regional language variants like European Portuguese (PT-PT) is heavily bottlenecked by dialectal overlap (mainly with PT-BR) and data processing scale. This paper presents an efficient pipeline to curate a production-ready PT-PT corpus from the Portuguese Web, spanning 411 TB of raw data from Arquivo.pt. We introduce a novel post-scraping block that removes boilerplate and line duplicates prior to filtering. This early-stage intervention increases final document yield by 19.04% by rescuing valid text that standard heuristic filters prematurely discard. Integrated with rigorous language identification, weighted fuzzy deduplication, and neural quality classification, our pipeline offers a scalable framework and a clean, representative corpus optimized for LLM pre-training.

cs.CL↗

arXiv2026

AMALIA Technical Report: A Fully Open Source Large Language Model for European Portuguese

Despite rapid progress in open large language models (LLMs), European Portuguese (pt-PT) remains underrepresented in both training data and native evaluation, with machine-translated benchmarks likely missing the variant's linguistic and cultural nuances. We introduce AMALIA, a fully open LLM that prioritizes pt-PT by using more high-quality pt-PT data during both the mid- and post-training stages. To evaluate pt-PT more faithfully, we release a suite of pt-PT benchmarks that includes translated standard tasks and four new datasets targeting pt-PT generation, linguistic competence, and pt-PT/pt-BR bias. Experiments show that AMALIA matches strong baselines on translated benchmarks while substantially improving performance on pt-PT-specific evaluations, supporting the case for targeted training and native benchmarking for European Portuguese.

cs.CL↗

arXiv2026

Dynamics of composition operators induced by odometers

We study the linear dynamics of composition operators induced by measurable transformations on finite measure spaces, with particular emphasis on operators induced by odometers. Our first main result shows that, on a finite measure space, supercyclicity of a composition operator implies hypercyclicity. This phenomenon has no analogue in several classical settings and highlights a rigidity specific to the finite-measure context. We then focus on composition operators induced by odometers and show that many dynamical properties that are distinct for weighted backward shifts collapse in this setting. In particular, for such operators, supercyclicity, Li-Yorke chaos, hypercyclicity, weak mixing, and Devaney chaos are all equivalent. In contrast to this collapse, we show that the classical equivalence between Devaney chaos and the Frequent Hypercyclicity Criterion for weighted backward shifts fails for odometers. Specifically, we construct a mixing, chaotic, and distributionally chaotic composition operator that does not satisfy the Frequent Hypercyclicity Criterion. This combination of rigidity and separation demonstrates that the dynamical behavior of composition operators induced by odometers differs sharply from that of weighted backward shifts.

math.DS↗

arXiv2024

Kitai's Criterion for Composition Operators

We present a general and natural framework to study the dynamics of composition operators on spaces of measurable functions, in which we then reconsider the characterizations for hypercyclic and mixing composition operators obtained by Bayart, Darji and Pires. We show that the notions of hypercyclicity and weak mixing coincide in this context and, if the system is dissipative, the recurrent composition operators agree with the hypercyclic ones. We also give a characterization for invertible composition operators satisfying Kitai's Criterion, and we construct an example of a mixing composition operator not satisfying Kitai's Criterion. For invertible dissipative systems with bounded distortion we show that composition operators satisfying Kitai's Criterion coincide with the mixing operators.

math.FA↗

arXiv2019

MementoMap Framework for Flexible and Adaptive Web Archive Profiling

In this work we propose MementoMap, a flexible and adaptive framework to efficiently summarize holdings of a web archive. We described a simple, yet extensible, file format suitable for MementoMap. We used the complete index of the Arquivo.pt comprising 5B mementos (archived web pages/files) to understand the nature and shape of its holdings. We generated MementoMaps with varying amount of detail from its HTML pages that have an HTTP status code of 200 OK. Additionally, we designed a single-pass, memory-efficient, and parallelization-friendly algorithm to compact a large MementoMap into a small one and an in-file binary search method for efficient lookup. We analyzed more than three years of MemGator (a Memento aggregator) logs to understand the response behavior of 14 public web archives. We evaluated MementoMaps by measuring their Accuracy using 3.3M unique URIs from MemGator logs. We found that a MementoMap of less than 1.5% Relative Cost (as compared to the comprehensive listing of all the unique original URIs) can correctly identify the presence or absence of 60% of the lookup URIs in the corresponding archive while maintaining 100% Recall (i.e., zero false negatives).

cs.DL↗