A sparse MoE NMT model that can be pruned into deployable sub-networks after training would change how multilingual systems are packaged for constrained devices, and EDWARD-NMT is built around that idea with adaptive routing and dual consistency. Alongside it, a tutorial on RLHF, SFT and DPO lays out the mathematical links between preference learning and inverse reinforcement learning, while ParaTaxe describes open-source tooling for aligning, processing and publishing multi-parallel corpora for digital-humanities projects.

Aligning Large Language Models with Human Feedback: Mathematical Foundations and Algorithm Design

If you are aligning MT or LLM outputs with preference data, the useful takeaway is the clean map from SFT to RLHF to DPO and the note that discrete-choice and IRL assumptions matter for how you set the objective.

This article provides an introduction to the mathematical foundations and algorithmic frameworks used to align Large Language Models (LLMs) with human intentions, preferences, and values. We discuss standard alignment techniques, such as fine-tuning (SFT), reinforcement learning with human feedback (RLHF), and direct preference optimization (DPO). We also explore the theoretical underpinnings of learning from human preferences, drawing connections to inverse reinforcement learning (IRL) and discrete choice models. We present state-of-the-art algorithms in a tutorial style, discuss their advantages and limitations, and offer insights into practical implementation. Our exposition is intended to serve as a comprehensive resource for researchers and practitioners, providing both a foundational understanding of alignment methodologies and a framework for developing more robust and scalable alignment techniques.

ParaTaxe: a Suite of Tools for Processing and Publishing Multi-parallel Corpora

For multilingual corpus work, ParaTaxe gives you an open-source pipeline plus web tools for aligning, managing, and publishing parallel collections, which matters because many academic corpora vanish without maintenance.

Over the past fifteen years, digital tools for parallelizing and aligning textual corpora have developed significantly within the field of digital humanities. Several projects and software tools have been created to compare different translations or versions of texts. Despite these technical advances, many parallel corpora produced in academic contexts quickly become inaccessible online due to the lack of maintenance or funding. To address this gap, we present a suite of open-source tools developed within the ACR Grimm Tradalign and ParaTaxe projects. This suite includes the AIlign aligner, based on multilingual embeddings, a Python processing pipeline for converting and aligning corpora, a web interface called ParaTaxe for managing collections of parallel corpora, and a web interface for exploring aligned corpora built with TEI-Publisher.

EDWARD-NMT: Elastic Dimensions With Adaptive Routing and Dual-consistency for Neural Machine Translation

For low-memory NMT deployment, EDWARD-NMT combines elastic dimensions, adaptive routing, and a dual-consistency loss so one checkpoint yields sub-networks that beat same-size dense Transformers without retraining.

Sparse Mixture-of-Experts (MoE) is a powerful paradigm for scaling Neural Machine Translation (NMT) and managing language heterogeneity. However, practical limitations hinder its deployment on resource-constrained devices: the entire expert pool must remain in memory, and the per-token compute budget is statically fixed during training by rigid top-k routing. We introduce EDWARD-NMT (Elastic Dimensions With Adaptive Routing and Dual-consistency for NMT), a novel MoE encoder-decoder architecture designed for a train once, deploy anywhere paradigm that exposes a family of deployable sub-networks at inference time. EDWARD-NMT integrates three key innovations: (i) Matryoshka-style elastic dimensions to flexibly reduce the memory footprint; (ii) an adaptive sigmoid router with a learned, task-conditioned threshold, allowing each token to dynamically determine its active expert count; and (iii) a dual-consistency objective to maximize the performance of the extracted sub-networks. Extensive experiments on PhoMT (En-Vi), WMT-14 (En-De), multilingual OPUS-16, and a multi-domain De-En benchmark validate our approach. The full-capacity architecture matches or exceeds strong MoE baselines with improved parameter efficiency. Furthermore, the sub-networks seamlessly extracted from a single checkpoint consistently outperform comparably sized dense Transformers without any retraining.

Papers announced by arXiv on Sunday, 25 October 2026, covering submissions from 22 October to 23 October 2026.