Incomplete relevance judgments can make abstention look better or worse than it is: on RECOR track 2, the shortcut that treats any context without gold documents as unanswerable shifts the evaluation enough to change conclusions across retrievers ranging from nDCG@10 .18 to .59 and three Qwen3 readers. Beyond that RAG result, one paper defines the Knowledge Layer Model for evidence-visible AI output and routing diagnostics, another introduces SHAKTI and TRIDENT for Hindi AI-text detection, a third studies CEFR level prediction across languages and text granularities, and a final case study follows a translator workstation migration from Windows to Linux as a test of digital sovereignty in professional translation workflows.

Better Retrievers, Worse Labels: How Incomplete Relevance Judgments Distort Abstention Evaluation in Conversational RAG

Incomplete relevance labels can make abstention look far worse than it is, because stronger retrievers surface many answer-bearing “no-gold” contexts, so evaluate sufficiency against the actual context, not the judgment shortcut.

Conversational RAG benchmarks such as MTRAG and RETECO evaluate retrieval and generation end to end: each system answers from its own retrieval, so abstention must be scored against whether each retrieved context is sufficient. A convenient shortcut reads sufficiency off the relevance judgments and treats a context without gold documents as unanswerable. We re-run the official baselines exactly and test this shortcut on RECOR (RETECO track~2) with four retrievers (nDCG@10 from .18 to .59) and three Qwen3 readers. Two open LLM judges and a panel of four small open models label each context by content, and a model of another family audits 250 contexts blind, with adjudication. The stronger the retriever, the larger the share of no-gold contexts that still contain the answer (28% to 72%) and the lower the shortcut's precision (.68 to .10; .70 to .21 under the audit). A reader agrees: Qwen3-4B answers 62--67% of these false misses correctly, against 23--30% of no-gold contexts judged insufficient and 1% with no documents. In most false misses, the answer comes from a hard negative mined for the same question, another turn's gold or a neighbouring chunk of the gold page. On MTRAG, whose judgments come from annotators, precision drops from .72 to .50. Under the shortcut, answerability AUROC is up to .19 too low, and the best sufficiency detector appears to be a retrieval score that never reads the documents. We release code, runs and labels, and recommend scoring refusals against content labels, not the presence of gold.

The Knowledge Layer Model (KLM): Evidence-Visible AI Output, Routing Diagnostics, and Negative Results

KLM is a record format for keeping evidence, uncertainty, and provenance visible, and the practical lesson is to preserve the text path end to end because packing errors can drop gold evidence in 118/221 cases.

AI answers require more than apparent correctness: reviewers need to inspect their recorded basis. We introduce the Knowledge Layer Model (KLM), an open, implementation-neutral specification for inference records with versioned epistemic assessments, evidence relations, and explicit unknowns. SEDIM, the knowledge and evidence architecture of Nage AI OS, is a reference implementation. We reference KLM v0.3 (repository tag v0.3.0). Empirical material consists of separate internal experiments, not end-to-end validation. A historical platform-domain routing benchmark reports 137/150 classifications across 15 classes (91.3%; nominal 95% Wilson interval 85.7-94.9%); the numerator is recovered from a rounded summary, and historical example membership is unverified. A native-model weight-path configuration yielded 0/15 successful answers in a corrected two-source test. A capacity-growth comparison could not establish a reliable effect because of a defective legacy scorer and missing original outputs. Lexical and scoped diagnostics each yielded AUC 0.008 for silent-versus-contradicted examples, below the registered minimum of 0.65. Legacy packing lost gold-containing text in 118/221 selected cases; repaired packing retained it in 221/221. These measure text presence, not answer quality. Python and TypeScript verifiers reportedly agreed on six checks per synthetic fixture; a separate production disclosure was not accepted. Restoration, re-derivation, human-rubric support, boundary isolation, and bounded-response evaluations remain proposed. Findings are internal and author-reported; historical observations are distinguished from later reproductions. Evidence visibility does not establish correctness, causal reliance, security, or regulatory compliance. Independent replication and a complete public audit package remain future work.

TRIDENT and SHAKTI: A Framework and Benchmark for Hindi AI-Text Detection

Round-trip translation can help detect AI text in Hindi, but the detector is sensitive to the translation model, so do not treat its scores as stable across engines or languages.

As Large Language Models (LLMs) become increasingly multilingual, distinguishing human-written from AI-generated text is essential for maintaining information integrity. However, research remains English-centric, with limited benchmarks for low-resource languages such as Hindi. To address this gap, we introduce SHAKTI, a large-scale Hindi benchmark covering four genres, eight LLM families, multiple prompting strategies, and adversarial transformations. We evaluate its quality through human annotation and analyze linguistic, grammatical, and sentiment-based differences between human- and AI-generated texts. Building on these findings, we propose TRIDENT, a round-trip translation-based detection framework with three variants: TRIDENT-S (static threshold), TRIDENT-D (dynamic threshold), and TRIDENT-C (logistic regression using multi-distance RTT features). TRIDENT leverages semantic consistency between original and round-trip translated texts to identify AI-generated content. Extensive experiments show that TRIDENT outperforms conventional and supervised baselines, with TRIDENT-S achieving the best overall performance. Further evaluations across seven languages, adversarial settings, and unseen LLMs demonstrate its multilingual applicability and robustness, while highlighting sensitivity to the translation model.

CEFR Level Assessment across Languages and Text Granularities

For CEFR tagging across languages and at sentence or document level, focal loss plus supervised contrastive learning handles class imbalance better than standard training and improves separation between adjacent levels.

Automatic assessment of textual CEFR levels is an important task in educational natural language processing. However, accurate CEFR level prediction remains challenging due to the naturally imbalanced distribution of CEFR levels and the ambiguous boundaries between adjacent levels. To address these challenges, we propose a unified training objective that combines focal loss with supervised contrastive learning. Focal loss mitigates class imbalance by emphasizing underrepresented and difficult examples, while supervised contrastive learning learns more discriminative representations by encouraging intra-class compactness and inter-class separation. We evaluate the proposed method on sentence- and document-level CEFR assessment in both monolingual and multilingual settings. Experimental results demonstrate that our method consistently outperforms strong baselines across all settings, establishing new state-of-the-art performance on multiple CEFR assessment benchmarks.

Digital sovereignty in professional translation workflows: a case study of migrating from Windows to Linux

A professional translation setup can run on Linux with local tools such as LoopCAT, LibreOffice, and TranslateGemma, but the real constraint is dependency control, not the operating system itself.

This study examines digital sovereignty in professional translation through a screening of computer-assisted translation tools and translation management systems, and a case study of migrating a translator's workstation from Windows to Linux. The framework combines three principles - Control, Continuity and Exit - with three practical requirements: native Linux operation, independence from mandatory external services, and independent access to essential data and resources. Drawing on documentary evidence and the limited number of responses received from providers, the ecosystem screening distinguishes Linux accessibility from operational independence: browser-based platforms may remain dependent on provider-controlled infrastructure. The migration case demonstrates the feasibility of reconstructing one professional workflow around a locally operable Linux core, using LoopCAT, LibreOffice and locally run TranslateGemma, alongside tools for optical character recognition, communication, credential management and backup. Selected cloud services remained available where needed. The findings suggest that digital sovereignty depends on how workflow dependencies are organized, rather than on individual applications or operating systems. It does not require technological isolation, but meaningful control over essential components, continuity of work and practical exit options. The framework is tailored to the case examined here and may require adaptation to other professional workflows.

Papers announced by arXiv on Sunday, 1 November 2026, covering submissions from 29 October to 30 October 2026.