Text Corpora

Retrieval-Augmented Generation (RAG) mitigates LLM hallucinations but introduces a critical vulnerability: corpus integrity. We present SilentRetrieval, a two-stage data poisoning attack that hijacks RAG systems through adversarially crafted yet fluent documents. Stage 1 uses Coordinated Beam Search, a multi-token joint optimization method with a fluency-similarity objective, to keep a poisoned host document retrievable while constraining perplexity. Stage 2 uses Context-Adaptive Trigger Generation, a lightweight trigger-fusion step driven by a frozen LLM, to integrate manipulation triggers into document content. Under a one-poisoned-document-per-query evaluation with synthetic target answers, SilentRetrieval achieves 84.6%/81.3% HR@10 and 57.5%/54.8% ASR-LLM on Natural Questions and MS MARCO, while maintaining near-benign perplexity. Cross-model evaluation across four target LLMs shows nontrivial effectiveness under a fixed trigger generator, and transfer tests against unseen retrievers, including ColBERT and commercial embedding models, yield 64.7% average HR@10 under the same injected-corpus protocol. In a sampled Wikipedia-scale evaluation, SilentRetrieval retains 74.2% HR@10 at a 0.016% poisoning ratio. Combined retrieval-side and generation-side defenses reduce attack success substantially but incur a latency trade-off. Human evaluation shows substantially lower flag rates than disfluent baselines, while remaining numerically more suspicious than benign content at the current sample size.

Time series forecasting in real-world settings often depends not only on historical observations, but also on external context that must be actively discovered from noisy, heterogeneous information sources. Yet existing context-aided forecasting benchmarks typically assume that the supporting context is already provided, leaving open whether agents can identify it on their own. Therefore, we introduce Dr-CiK, a benchmark for evaluating whether agents can retrieve forecasting-relevant supporting context from a document corpus, filter out distractors, distill the retrieved context into forecast-useful evidence, and generate forecasts supported by that evidence. Through context ablations and evaluations of state-of-the-art deep research and forecasting methods paired together, we show that high-quality context substantially improves forecasting performance in Dr-CiK. However, most existing DR agents recover only a small fraction of the ground-truth supporting evidence (usually <5%), are frequently misled by distractors (>80% distractor citations), and can cause forecasters to perform worse with retrieved context than without context. Our results motivate research on foresight-driven agents that search for the right context to predict the future.

Most Automatic Speech Recognition (ASR) systems formulate transcription as a prediction problem over orthographic units such as characters, subwords, or words. Although effective, such representations do not explicitly reflect the phonetic structure of speech and often require large vocabularies to maintain adequate coverage. In this work, we are motivated from the phonemic features of Vietnamese to propose a Syllabic-Structure Decoder for ASR, which models speech at the phoneme level instead of the orthographic level. Our approach explicitly captures the phonological composition of syllables, enabling the decoder to generate valid syllabic structures from a compact phonemic inventory. This design more closely aligns with the phonetic realization of speech while significantly reducing vocabulary size. Experimental results on two benchmarks: LSVSC, representing standard speech, and UIT-ViMD, a multi-dialect corpus containing diverse regional pronunciations, show that our method consistently outperforms strong previous baselines, especially pretrained baselines such as PhoWhisper and Wav2Vec2, despite using a substantially smaller vocabulary and no additional training resources. These results highlight the effectiveness of phoneme-based syllabic modeling for ASR in this language. Code for experimental reproducibility will be publicly available upon the acceptance of this paper.

Data-aware post-training quantization (PTQ) minimizes a per-token reconstruction loss on a small calibration corpus, implicitly weighting positions by their empirical frequency. For \textbf{A}utomatic \textbf{S}peech \textbf{R}ecognition (ASR), this misaligns with tail-sensitive risk: names, numerals, and domain-specific words receive proportionally little calibration mass. We propose \textbf{Tail-Aware Reconstruction Quantization} (\TARQ), a label-free PTQ framework that shifts calibration toward the lexical tail via \textbf{\rareBAL}, a closed-form per-Linear-layer rule equalizing common/tail mass, paired with a metric-consistent residual correction. \TARQ\ requires no entity labels, no curated calibration set, no validation decoding, and no additional training. Across eight ASR backbones and six datasets at W4G128, \TARQ\ improves mean rare-\textbf{W}ord \textbf{E}rror \textbf{R}ate (rare-WER) without an aggregate-WER regression, achieves the lowest cross-corpus rare-WER swing among compared methods, and transfers to entity-rich benchmarks (ProfASR, ContextASR-Speech-En) without entity supervision.

Modern Large Language Models (LLMs) are often criticized for producing repetitive and homogeneous text, despite possessing vast latent vocabularies. While previous research has focused on model knowledge and training data, we investigate the role of decoding mechanics in suppressing linguistic diversity. We introduce the Word Coverage Score (WCS), a metric that quantifies the extent to which contextually appropriate human vocabulary is mathematically pruned by standard sampling filters (e.g., Top-$p$, Top-$k$, and Min-$p$). Rather than assessing static knowledge, the WCS measures the lexical survival rate of low-frequency, high-information human words as a function of sampling parameters. By auditing open-weight models on human-authored corpus fragments, we identify which logical lexical choices are rendered unreachable by the decoder, even when they reside within the probability space. Our results provide quantitative evidence that industry-standard sampling defaults act as unintended censorship mechanisms, smoothing the unique textures of human expression into a homogenized discourse. The WCS offers a rigorous framework for optimizing the trade-off between text coherence and lexical richness, providing a diagnostic tool for preserving the diversity of human language in generative models.

Retrieval-Augmented Generation (RAG) systems for question answering typically retrieve evidence by semantic similarity between the query and document chunks. While effective for unstructured text, this approach is less reliable on semi-structured corpora where answering may require exact filtering, aggregation, or exhaustive retrieval over structured attributes across multiple documents. Symbolic approaches support such operations, but they are often brittle on noisy natural-language corpora. We address this gap with DualGraph, a RAG framework that represents documents through two complementary views: a Textual Knowledge Graph for semantic retrieval and a Symbolic Knowledge Graph for symbolic querying over typed subject--predicate--object triples. Building on these two components, we provide multiple strategies for selecting or combining semantic and symbolic evidence.We also introduce SpecsQA, a benchmark from a commercial shopping website with semi-structured product documents and manually curated questions spanning open-ended and specification-oriented retrieval. Experiments show that DualGraph consistently outperforms state-of-the-art dense-retrieval, GraphRAG, symbolic, and table-oriented baselines across question types.Code and data are available at https://github.com/corneliocristina/DualGraphRAG.

State-of-the-art performance for Automatic Speech Recognition (ASR) largely depends on the availability of large-scale labeled corpora. This creates a demand for increased data collection efforts, particularly for under-represented languages and dialectal varieties. Due to having considerably fewer speakers (around 11 million), European Portuguese (EP) is overshadowed by Brazilian Portuguese (BP) (around 200 million speakers) in currently available large-scale speech data resources, resulting in under-performing speech-based systems for EP users. To address this gap, and following similar data collection efforts for other languages, we present FalAR, a large-scale, speaker-annotated speech corpus of European Portuguese parliamentary sessions. Spanning approximately 20 years, FalAR comprises 5,800 hours of speech data. In addition, 4,850 hours have speaker identity annotations, for a total of 1,180 speakers with associated metadata including age, gender, political affiliation, and parliamentary role. The corpus was built using a state-of-the-art EP CAMÕES ASR model for transcription-reference alignment. In this paper, we describe the data collection process, together with the main characteristics of the FalAR corpus. Furthermore, we evaluate the trade-off between data quantity and alignment accuracy on ASR performance, with our experiments demonstrating that incorporating FalAR as pre-training data yields up to 14% relative WER improvement over baseline models.

We present BhashaSetu, a linguistically enriched English--Marathi parallel dataset addressing persistent data limitations in low-resource neural machine translation (NMT). Marathi, spoken by over 95 million people, remains underrepresented in high-quality parallel corpora across diverse domains. Our dataset comprises 2.78 million sentence pairs from heterogeneous sources including news, politics, healthcare, literature, and culture, with stemmed and lemmatized representations to support morphology-aware analysis. We benchmark multiple state-of-the-art translation models using BLEU, spBLEU, chrF++, and TER metrics, and conduct parameter-efficient fine-tuning of NLLB-200-distilled-600M using LoRA. A key finding from our ablation: corpus-level deduplication is the single largest preprocessing contributor to downstream quality (removing it reduces performance by 1.17 BLEU and 2.21 chrF++), demonstrating that disciplined cross-source corpus hygiene is a low-cost, high-impact intervention for low-resource, morphologically rich languages. The dataset is publicly released to promote reproducible and linguistically informed low-resource NMT research.

This paper presents a multilingual customer service self-help corpus comprising 1,122 manually validated documents in Finnish, Danish, Norwegian, and Swedish, totaling 274,599 words and 1,884,833 characters. The documents have been sourced from the public self-help pages of four Nordic telecommunications operators and subsequently filtered for person-identifiable information and relevance through a combined LLM and human annotation pipeline. Domain-specific datasets for Nordic languages remain scarce, particularly in customer service: a domain of growing importance for retrieval-augmented generation, cross-lingual transfer learning, and emerging agent-based service architectures. An analysis of the corpus reveals substantial variation in document length and structure across operators, reflecting distinct editorial strategies, as well as broad topical coverage spanning network hardware, mobile services, TV and streaming, billing, and account management. The dataset is publicly available under a CC-BY-NC-SA-4.0 license at https://zenodo.org/records/20732652, intended to support reproducible research in Nordic NLP and information retrieval.

While previous research in multivariate time series forecasting has focused on developing complex holistic models, this work advocates for a shift toward a granular, component-level understanding of their impacts. We propose TSCOMP, the first large-scale benchmark that systematically deconstructs deep forecasting methods into their core, fine-grained components--spanning series preprocessing, encoding strategies, network architectures including specific and large time-series models, and optimization methods. Using constrained orthogonal experimental design and extensive evaluations, we conduct multi-view analyses that reveal component effectiveness across different backbones, data characteristics, and their interactions. Beyond providing insights, this benchmark establishes a fine-grained performance corpus comprising over 20,000 model-dataset evaluations, which supports the learning of automated component selection, enabling zero-shot model construction on new datasets. Our experiments demonstrate that the corpus-driven approach, despite its simplicity, consistently outperforms state-of-the-art methods, validating the soundness of our evaluation design and confirming that systematic component selection surpasses manually designed complex architectures. All code and the performance corpus are publicly available at https://github.com/SUFE-AILAB/TSCOMP.

Objective. Outpatient notes carry follow-up instructions pairing actions with future times ("MRI brain in two weeks"). Extracting (action, date) pairs supports scheduling and audit, but generative extractors miss the date because linking and arithmetic are implicit in decoding. We test a hybrid neural-symbolic pipeline against direct generation. Methods. We define TestSpecification and TimeSpecification entities and a ScheduledFor relation. BioBERT feeds BIO tagging and a biaffine linker; entities are canonicalized via a 28-action ontology and times normalized to day offsets deterministically. We evaluate on a 2,000-note synthetic outpatient corpus with action-disjoint splits (18 train, 6 OOV-test) against zero-shot GPT-4o-mini and LoRA-fine-tuned LLaMA-3 8B with note-level bootstrap 95% CIs. Results. On 259-note seen and OOV splits the hybrid pipeline achieves Test-Time Pair F1 of 0.997 and 0.986 with 0.00-day MAE. Baselines reach high action F1 (LLaMA-3 0.992; GPT-4o-mini 0.963 seen) but Pair F1 stays at 0.51-0.57 (LLaMA-3) and 0.53 (GPT-4o-mini), CIs non-overlapping with the hybrid. Conclusion. Separating learned entity extraction from deterministic date arithmetic outperforms generation on this benchmark, generalizes to held-out actions, and exposes failure modes. Transfer to real EHR notes is the next validation; a first-pass realism check is in Limitations.

Automated Essay Scoring (AES) for English proficiency assessment increasingly relies on pretrained transformer models, yet these models are typically trained on general-domain English and may under-represent second-language learner writing. This study investigates whether domain-adaptive continued pretraining (DAPT) on a learner-writing corpus improves transformer-based AES for English proficiency assessment. We perform DAPT on BERT, RoBERTa, and DistilBERT using the EFCAMDAT corpus, then compare the adapted models with their original checkpoints on two English proficiency test datasets, FCE and IELTS, in both in-domain scoring and few-shot cross-dataset transfer. Full-corpus DAPT produces mixed effects across models, datasets, and metrics. Subsequent lexical and syntactic analyses suggest mismatches between EFCAMDAT and the downstream datasets in proficiency level, genre, and communicative purpose. We therefore repeat DAPT using proficiency-specific EFCAMDAT subsets across all three encoder architectures. Proficiency-specific DAPT frequently outperforms full-corpus DAPT and, in some settings, even the non-adapted baseline. Overall, continued pretraining on learner writing can improve in-domain AES, but its benefits depend on both the proficiency composition of the pretraining data and the underlying encoder architecture, and do not consistently extend to cross-test transfer.

Evidence construction--the stage that determines which passages reach the language model before generation begins--is evaluated paradigm by paradigm, leaving practitioners with no principled way to diagnose which organization strategy fails, where, or why. We introduce AuthTrace, a diagnostic benchmark built on thematically dense single-author corpora where near-miss distractors share style, topic, and vocabulary with the required evidence. AuthTrace provides explicit quoted evidence, exact fan-in annotation, and a unified pack-level protocol measuring evidence recall, evidence precision, and answer correctness. A fan-in gradient--the number of source documents required to support the answer--serves as the primary diagnostic axis, enabling controlled comparison across retrieval, memory, graph, and structured-evidence paradigms. Evaluating eight systems across two QA models, we find that evidence recall is the strongest observed predictor of answer correctness under the primary reader-judge pair (r = 0.96); most failures stem from missing evidence rather than answer synthesis. Fan-in further exposes paradigm-specific collapse patterns: flat retrieval degrades 2-3x faster than thematically organized evidence construction. These results show fan-in decomposition to be a reusable diagnostic lens for identifying where evidence-construction systems fail and which paradigm best serves a given workload.

AI-associated lexical shifts have been documented mainly in Scientific English. We extend this work to 34 languages in the WMT News Crawl corpus, refining a split-halves continuation diagnostic that compares GPT-4.1 continuations with matched human gold-standard text. For each language, we derive ranked AI-overused lemmas using log prevalence ratios. We find substantial cross-lingual semantic convergence: semantically related concepts recur across typologically diverse languages, with 'emphasize'-type verbs appearing in 24 of 34 languages. Embedding-based and manual analyses support this pattern. We also examine diachronic uptake in news writing before and after ChatGPT's release. Tracking each language's top 20 AI-overused items, we find prevalence increases in 26 of 34 languages from 2020-2021 to 2023-2024, with a mean change of +15.1%, whilst matched baseline words show no comparable increase (-4.5%). In 10 languages with longer historical coverage, longitudinal analyses show post-2022 increases that exceed the modest shifts observed in earlier periods, though with smaller effect sizes than in Scientific English. We validate our approach extensively, including across seeds, model variants, data sizes, model families, and more. Our findings are consistent with the view that AI-associated lexical preferences extend beyond English and may exert cross-lingual homogenising pressure on global language use.

Time series foundation models (TSFMs) are increasingly pretrained on large corpora, raising concerns that evaluation datasets may have been exposed during pretraining and thus yield overly optimistic performance estimates. Auditing such contamination is challenging in time series because signals are continuous and heterogeneous, and often lack corpus documentation. To the best of our knowledge, this is the first work to study pretraining contamination auditing for TSFMs. We formalize the problem of pretraining contamination auditing for TSFMs and propose TSFMAudit, a method based on probe adaptation dynamics. Our key intuition is that contamination manifests as unusually efficient adaptation: after a fine tuning probe, contaminated datasets tend to exhibit faster loss reduction with smaller backbone movement. We evaluate TSFMAudit on 6 TSFMs and 187 datasets using documented training source evidence as supervision, and compare against 10 competitive baselines adapted from the LLM literature.

Long-term conversational memory is a retrieval workload classical IR was not built for: the index grows during the query stream, query types shift intra-session, and the latency budget per retrieval is sub-10 ms. Lucene-class engines treat the index as static and the query as stateless, leaving the workload's structure unexploited. AgentIR treats fusion as a per-query decision along two axes: which fusion to apply (BM25, Dense, RRF, or agent-aware RRF), and whether the ~52 ms dense channel is worth running at all. The second axis is a confidence-triggered cascade router that decides from the BM25 top-k margin alone and re-tunes across workloads without retraining. On LongMemEval (n=500), where the dense channel does add information, the cascade skips 63% of queries at parity LLM-judged accuracy (2.67x faster under two judges, paired bootstrap p>=0.88); per-qtype thresholds extend this to 5.76x under 5-fold cross-validation. On LoCoMo (n=1,982), where BM25 alone is already the strongest single system, the same trigger auto-tunes to a 100% skip rate (132x faster, +0.089 Hit@5). Capacity on a shared 8-core VM rises from ~154 to ~1,400 concurrent agents (9x). Underneath the cascade, a time-partitioned index does O(log 1/epsilon) work independent of corpus size: 1234x corpus growth costs only 3.6x latency, ending in 1769x over sequential at sub-100 us p50 on 5M records. At parity quality with Lucene on 9 BEIR datasets up to 8.8M docs, the substrate runs 10x geo-mean over Pyserini 8T and 11x over PISA-1T BlockMax-WAND; an A100 reaches 1.8-39x over Pyserini 8T; chunked index build sustains 56.8K docs/sec on MS MARCO. Three subtle BM25/GPU correctness pitfalls that silently regress nDCG@10 by 6-8x are documented and fixed; post-fix CPU and GPU agree within 0.0002 nDCG@10 on all eight datasets that fit a single A100.

Large Language Models (LLMs) have shown great promise in multilingual machine translation (MT), even with limited bilingual supervision. However, fine-tuning LLMs with parallel corpora presents major challenges, namely parameter interference. To address these issues, we propose Mix-MoE, a mixed Mixture-of-Experts framework designed to train LLMs for multilingual MT. Our framework operates in two distinct stages: (1) post-pretraining with MoE on monolingual corpora, and (2) post-pretraining with MoE on parallel corpora. Crucially, we divide the MoE layers into two specialized groups: Language Model Experts (LM Experts) and Machine Translation Experts (MT Experts). LM Experts are designed to capture and retain the monolingual knowledge learned by the pre-trained LLM. MT Experts, on the other hand, are specifically trained to acquire and store bilingual translation knowledge. Furthermore, to facilitate effective interaction between these specialized experts and leverage potential underlying structural patterns in text, we introduce a routing mechanism enhanced by Fourier Transform features derived from model representations. The experimental results demonstrate that Mix-MoE excels in multilingual MT, significantly outperforming existing baselines and showing notable progress in mitigating parameter interference.

Aiming at identifying unexpected inputs from unknown classes, out-of-distribution (OOD) detection has emerged as a pivotal approach to enhancing the reliability of machine learning models. This paper focuses on the burgeoning paradigm of post-hoc OOD detection with pre-trained vision-language models (VLMs), where a popular pipeline is to detect OOD inputs by examining their affinities between ID labels and negative labels, i.e., those semantically different from ID labels. Due to the unavailability of target OOD labels, existing works predominantly rely on heuristic rules to mine negative labels from unlabeled wild corpus data. Despite the empirical success, we argue that the power of VLM-based OOD detection has yet to be fully unleashed since the notorious false negative problem is far from addressed in the literature. With this motivation, we are interested in addressing the challenge of mining true negative labels for OOD scoring. To this end, we develop a theoretical framework for correcting the sampling bias of negatives labels by indirectly approximating the distribution of negative labels. Perhaps surprisingly, we show that the debiased negative mining can be naturally converted into Monte-Carlo sampling based on ID labels and the unlabeled wild corpus data. Extensive experiments empirically manifest that our method establishes a new state-of-the-art in a variety of OOD detection setups. Code is publicly available at \href{https://github.com/60pen9/Debiased-Negative-Mining-Improves-OOD-Detection-with-Pre-trained-VLMs}{\textcolor{red}{here}}.

Semantic type mismatch between a noun and its context is central to coercion phenomena. This paper introduces a graph-based method to examine how lexical and contextual type information is reflected in word embeddings. We select nouns from ten semantic types, annotate corpus instances for type matching (matching vs. coercion vs. other mismatch vs. unrestricted), and construct graphs using BERT and sense-enhanced embeddings. Two metrics -- Neighbor Type Probability (NTP) and Neighbor Type Entropy (NTE) -- are proposed to analyze neighborhood type distributions. Results show that graphs constructed with sense-enhanced embeddings reflect semantic type information better, and matching and mismatch sentences can be distinguished through the proposed metrics.

While factual correctness and task-performance have been in focus of Large Language Model (LLM) research for a long time, the fundamental question of how human-like generated texts are on a linguistic level has been underexplored. From a corpus-linguistic perspective, language production is inherently context-dependent, with distinct communicative contexts giving rise to differences in frequencies and co-occurrence patterns of linguistic features. A text failing to adhere to these patterns can be content-wise correct, but still be unfavorable to human readers. In this work, we propose a context-aware evaluation framework in which human-likeness is assessed using a two-sample problem between the linguistic feature distribution of a human reference corpus for a given register and a corresponding LLM-generated corpus. We implement this framework using the Maximum Mean Discrepancy (MMD) and the 67 lexico-grammatical features introduced by Biber, which are commonly applied in corpus linguistics. In our experiments, we compare seven instruction-tuned, open-source models across five English-language datasets spanning distinct registers against a human baseline. While across all tested setups, LLMs deviate from the human baseline, which models are closest to human language depends on the register and is not dictated by model size.

We benchmark Google Embeddings (GE2), a Vertex-AI-hosted bi-encoder with 2,048-token context and explicit task-type conditioning, against five open-source alternatives: BGE-M3, E5-large, Multilingual-E5-large (mE5-L), LaBSE, and Paraphrase-Multilingual-MPNet (mMPNet). Evaluation covers four BEIR subsets, a synthetic Italian RAG corpus, a chunking ablation considering 5 sizes of tokens with three strategies, and per-query latency on commodity CPU hardware. GE2 ranks first on every task, achieving BEIR avg.nDCG@10 = 0.638 and IT-RAG-Bench nDCG@10 = 0.282, but at 231.6 ms median latency, it is roughly 14x slower than the fastest local models. mE5-L reaches within 0.003 nDCG of GE2 on Italian at 31 ms, making it the preferred option when sub-100 ms SLAs matter. A more striking finding concerns LaBSE, which, despite widespread multilingual deployment scores 0.188 average nDCG@10 on BEIR, below every dedicated retrieval model including mMPNet. Chunking experiments show that all six models saturate at 32-token chunks on our corpus, with semantic chunking providing measurable gains only at 16 tokens.

Language models trained on observed sequences are often described as learning the conditional distribution of the next token given previous tokens. This description is only conditionally correct. A model trained on realized token trajectories does not observe full conditional laws; it receives sampled continuations. Moreover, real language generation is conditioned not only on previous words but also on non-textual circumstances: facts, events, intentions, goals, beliefs, social context, and task-specific constraints. This paper distinguishes three objects that are often conflated: the full conditional language process conditioned on latent circumstances, the marginal text-only process obtained by integrating those circumstances out, and the model-induced distribution learned from finite observed corpora. The paper argues that interpreting model training as estimating the marginal text-only law requires strong assumptions of stationarity, representativeness, and ergodicity, assumptions that are standard in statistical estimation but problematic when applied to heterogeneous language corpora. Even if these assumptions hold, the marginal text-only law is useful only when the observed prefix is an approximately sufficient statistic for the latent circumstances relevant to continuation. In information-theoretic terms, usefulness requires that the residual conditional mutual information between the next token and the omitted circumstances, given the observed text, be small. The paper then extends this argument to heterogeneous training corpora. Finally, the paper interprets Retrieval Augmented Generation (RAG) and tool use as conditional sufficiency devices.

Generative text-to-video systems can produce visually fluent educational clips, but they rarely encode the pedagogical content knowledge (PCK) needed for effective instruction, including prerequisite-aware sequencing, learner-adaptive depth, and sustained cognitive engagement. We present CourseBlueprint, a course-grounded pipeline for adaptive pedagogical video generation. Given a topic and learner persona, the system generates a structured teaching blueprint in a single forward pass over an undergraduate biomedical-imaging corpus (BMED 2300; twenty-three lectures, 1,116 slides). Instead of ad-hoc prompt chaining, the pipeline uses typed intermediate representations with validation: a scaffolding module builds a stage-labeled prerequisite concept graph with deterministic cycle removal, an adaptive controller assigns per-concept style specifications, and an engagement generator produces narration following a fixed hook->retrieval->core->analogy->forward contract. A deterministic slide-image override further grounds the rendered video by reusing instructor slides whenever retrieval confidence is high. We also release a reusable benchmark corpus and an evaluation harness combining repeated LLM-judge scoring with regex-grounded objective metrics. In a five-topic ablation, removing the engagement contract reduces the engagement score from 5.00 to 1.20, the adaptive score from 4.80 to 3.40, Flesch readability from 38.0 to 19.8, and analogy and retrieval-prompt counts to near zero. The slide-image override converts a 0/9 corpus-grounding failure into 9/10 successful slide matches on the same topic. These results show that pedagogical video quality depends less on surface fluency than on explicit, typed instructional contracts that make scaffolding, adaptation, engagement, and grounding auditable.

We present Epicure, a family of three sibling skip-gram ingredient embeddings retrained from scratch on a multilingual recipe corpus. We aggregate 4.14M recipes from 11 sources spanning seven languages, English, Chinese, Russian, Vietnamese, Spanish, Turkish, Indonesian, German, and Indian-English, and normalise the raw ingredient strings to 1,790 canonical entries via an LLM-augmented pipeline. A 203,508-edge ingredient-ingredient NPMI graph and an 80,019-edge typed FlavorDB ingredient-compound graph, 2,247 typed compound nodes across 15 categories, seed three Metapath2Vec variants that share architecture and hyperparameters and differ only in the random-walk schema: Cooc walks the co-occurrence graph only, Chem walks the typed compound metapaths only, and Core blends both via injected ingredient-ingredient walks at controlled mixing, placing each model at a distinct point on the chemistry-vs-recipe-context spectrum.

This paper presents the Arabic Women and Society Corpus, a ten year collection of 252,487 public Arabic Facebook posts related to women's empowerment and social wellbeing. The corpus was collected from 51,660 pages across 77 countries between 2013 and 2024, resulting in more than 267 million user interactions. Each post includes engagement metrics such as shares, comments, and emotional reactions, providing a unique view of audience sentiment and social attention. The data were processed using an automated pipeline with language identification, normalization, and metadata cleaning to ensure reliability and reproducibility. The corpus enables large scale analysis of gender discourse, social reform, and emotional engagement across Arabic dialects. It supports research in Arabic natural language processing, computational social science, and digital communication studies. The dataset and accompanying documentation will be released under request for research use.

Classifier-based Quality Filtering has recently emerged as a fundamental technique in constructing pre-training corpora. The ability to deploy a single model that can replace or supplement a set of heuristics has proven effective across numerous Large Language Models. In this work, we expose a critical vulnerability in this approach by demonstrating how a straightforward Wikipedia-style reformatting operation can substantially alter a model's quality assessment and enable low-quality content to surpass filtering thresholds. Our analysis reveals that the FineWeb-Edu CQF model would reverse its filtering decision for approximately 7% of evaluated documents, thereby admitting content into the pre-training corpus that would otherwise have been excluded.

We present ArabDiscrim, a decade-long lexical resource and corpus of 293K public Arabic Facebook posts (2014--2024) discussing racism and discrimination. Unlike existing Twitter-centric datasets, ArabDiscrim integrates platform-native engagement signals, including reactions, shares, comments, and page metadata, enabling joint analysis of language and audience response. The resource includes 200 curated terms (100 racism-related and 100 discrimination-related) with morphological regex families (13+ inflections per lemma), and 20 discrimination axes capturing identity-based grounds for unequal treatment. It also provides explicit attribution patterns. Released under a restricted research-use license for ethical compliance with platform terms, ArabDiscrim supports weak supervision, axis-aware sampling, and platform ecology research. By bridging lexical depth and ecological validity, it establishes a foundation for fairness-oriented, platform-aware Arabic NLP.

Existing binary corpora typically capture only one or two axes of binary variation: they either provide cross-compiler builds without a temporal axis, or CVE labels for single-build binaries. None combine cross-build diversity, cross-version history, and CVE labels into a queryable structure. We present ASSEMBLAGE-DEEPHISTORY, which consolidates these dimensions into a unified framework where every binary's compilation context, source code, vulnerable functions, and package version are stored as first-class metadata. ASSEMBLAGE-DEEPHISTORY comprises 73,610 binaries spanning 248 open-source projects, compiled across GCC, Clang, and MSVC at multiple optimization levels on Linux and Windows, with multi-year historical builds. Each binary is indexed in a database that links it to its source code, functions, debug info, variant builds, historical versions, and vulnerable functions. Three analyses demonstrate this structure's value: (1) a three-stage LLM benchmark (recognition, strategy-guided detection, and cross-build transfer) to test whether LLMs reason about binary vulnerabilities or pattern-match on build-specific artifacts; (2) a comparison of MalConv embeddings, jTrans function embeddings, and TLSH fuzzy hashes quantifying how same-package versions cluster in each space; and (3) a Bayesian regression decomposing binary similarity into contributions from temporal distance, file changes, and commits.

Diagnosing failures in LLM agents remains largely manual. Practitioners inspect a small subset of execution traces, form ad-hoc hypotheses, and iterate. This process misses patterns that only emerge across trace populations and does not scale to production corpora where individual traces span tens of thousands of tokens. We formalize the problem of corpus-level trace diagnostics. Given a corpus of execution traces, the goal is to produce grounded natural-language insights that characterize systematic behavioral patterns across trace groups, each linked to supporting evidence. We present the Insights Generator (IG), a multi-agent system that answers diagnostic questions by proposing and testing hypotheses across the trace corpus to produce an evidence-backed insights report. We evaluate IG across qualitative and objective dimensions, spanning rubric-based report assessment and downstream performance improvements achieved by implementing IG insights. Human experts using IG reports improve scaffold performance by 30.4pp over the unmodified baseline scaffold, and coding agents leveraging IG-derived insights show consistent and stable gains. Across benchmarks, IG's scout-investigator architecture produces findings comparable in detection coverage to competing approaches, while domain experts rated IG reports as leading depth and evidence quality.

This paper introduces JobArabi, a large-scale corpus of Arabic job announcements collected from social media between January 2024 and October 2025. The dataset contains 20,528 public posts from X and captures more than two years of employment-related discourse across Arabic-speaking online communities. The corpus was compiled using a linguistically informed query framework covering 21 Arabic keyword families that reflect gendered, plural, formal, and dialectal expressions of recruitment language. The resulting dataset includes posts from institutional, commercial, and individual accounts and provides metadata such as timestamps, engagement indicators, and geolocation when available, enabling temporal and regional analysis of employment discourse. Quantitative analysis reveals several sociolinguistic patterns in online recruitment, including the persistence of gendered hiring language, regional variation in occupational demand, and the emotional framing of recruitment messages. These findings highlight the potential of Arabic social media as a resource for studying labor market communication and linguistic change. The JobArabi corpus, together with documentation and collection scripts, will be released to support research in Arabic NLP, computational social science, and digital labor studies.

The increasing volume of scientific research necessitates effective communication across language barriers. Machine translation (MT) offers a promising solution for accessing international publications. However, the scientific domain presents unique challenges due to its specialized vocabulary and complex sentence structures. In this paper, we present the development of a collection of parallel and monolingual corpora for the scientific domain. The corpora target the language pairs Spanish-English, French-English, and Portuguese-English. For each language pair, we create a large general scientific corpus as well as four smaller corpora focused on the domains of: Cancer Research, Energy Research, Neuroscience, and Transportation research. To evaluate the quality of these corpora, we utilize them for fine-tuning general-purpose neural machine translation (NMT) systems. We provide details regarding the corpus creation process, the fine-tuning strategies employed, and we conclude with the evaluation results.

We present the University of Florida Gators submission to the AmericasNLP 2026 shared task on cultural image captioning for Indigenous languages. Our two-stage pipeline generates a Spanish intermediate caption with Qwen2.5-VL, then produces the target-language caption using retrieval-augmented many-shot prompting with Gemini 2.5 Flash. We achieve 164.1%, 131.7%, and 122.6% improvements over the shared task baseline for Bribri, Guaraní, and Orizaba Nahuatl captioning, respectively, in our dev set evaluation and maintain >150% improvements for the Bribri and Orizaba Nahuatl languages in the test set evaluation. We find retrieval is highly language-dependent, beneficial only for large, in-domain corpora, and that synthetic data augmentation accounts for around 28 chrF++ of the dev set Guaraní performance gain. Our submission is the overall winner of the shared task, placing second out of five finalist submissions in human evaluations of target-language captions.

Biomedical named entity recognition (NER) and entity linking (EL) strongly depend on annotated corpora, but the utility of these resources for benchmarking is often assumed rather than characterized. We present a corpus-centric framework for diagnosing benchmark-relevant properties directly from corpus annotations, concept links, train-test splits, document metadata, and terminology mappings. The framework organizes standardized statistics into five families: (1) scale, density and label distribution, (2) lexical and conceptual structure, (3) train-test overlap, (4) metadata composition, and (5) terminology coverage where applicable. Applying the framework to nine corpora spanning diseases, chemicals, and cell types, we find that corpus properties can differ substantially, even when they address the same apparent task. We find differences in the evaluation signal they provide, the generalization demands they impose, the degree of train-test reuse they permit, and the regions of biomedical literature and concept space they represent. These differences suggest that commonly reported corpus statistics can be insufficient to characterize what biomedical NER and EL benchmarks evaluate. We argue that corpus-centric diagnostics provide a practical framework for analyzing corpora beyond surface descriptors such as corpus size and entity type, for identifying potential transfer risks, and for interpreting the scope of benchmarking conclusions. We release the framework as open-source code with an interactive dashboard to support reproducing our analyses and characterizing additional corpora.

Has the style of scientific communication changed due to the growing use of large language models in the writing process? We address this question in the domain of Natural Language Processing by leveraging two data resources we create: a naturalistic corpus of over 37,000 papers from the ACL Anthology (2020-2024); and a synthetic dataset of 3,000 human-written passages and their LLM-generated improvements. We first implement a series of diachronic lexical analyses, showing that both word frequency and usage contexts have changed significantly over time, indicating semantic specialization in some cases and generalization in others. Broadening our perspective, we then model a range of more complex stylistic features and find that LLM-modified texts more frequently contain certain syntactic constructions, more complex and longer words and a lower lexical diversity. Finally, we connect these changes in writing practices to subjective reading experience through a pilot annotation study with 20 domain experts. They overall rate LLM-improved texts as more understandable and exciting, but also express negative qualitative attitudes towards LLMs, highlighting the strongly subjective effect of AI-assisted writing on reading experience.

This paper investigates chunking strategies for retrieval-augmented generation on German statutory law, using the German Civil Code as a structured benchmark corpus. We implement and compare a range of segmentation approaches, including structural units (sections, subsections, sentences, propositions), fixed-size windows, contextual chunking, semantic clustering, Lumber-style chunking, and RAPTOR-based hierarchical retrieval. All methods are evaluated on a legal question-answering dataset with section-level gold labels, measuring recall, query latency, index build time, and storage requirements. Results show that chunking strategies aligned with the inherent legal structure - particularly section and subsection - based retrieval-achieve the highest recall, while more complex approaches that override this structure perform worse. These simpler methods also offer favorable computational efficiency compared to LLM-intensive techniques such as contextual chunking, RAPTOR, and Lumber. The findings highlight a key trade-off between semantic enrichment and operational cost, and demonstrate that preserving domain-specific structure is critical for effective legal information retrieval.

Existing affective-computing, social-signal-processing, and meeting corpora capture important parts of human interaction, but they rarely support analysis of affect in co-located groups as a coupled individual, interpersonal, and group-level process. The required signals (per-participant physiology, eye movement, audio, self-report, task outcomes, and personality) are usually fragmented across separate dataset traditions. We introduce GroupAffect-4, a multimodal corpus of 40 participants in 10 four-person groups, each completing four ecologically varied collaborative tasks spanning information pooling, negotiation, idea generation, and a public-goods game. Each participant is instrumented with a wrist-worn physiology sensor, eye-tracking glasses, and a close-talk microphone; sessions include continuous affect self-reports, post-task questionnaires, task outcomes, and Big-Five personality scores, all time-aligned to a shared clock. The dataset covers over 91% of expected physiology windows and 98% of eye-tracking windows, with strong task validity confirmed by a clear affective manipulation check across the negotiation block. We define fifteen benchmarkable targets spanning three analysis levels -- within-person state, between-person traits, and group dynamics -- and report leave-one-group-out feasibility baselines establishing the dataset's evaluative scope. GroupAffect-4 is released with a BIDS-inspired structure, Croissant metadata, a datasheet, per-session quality reports, and open processing scripts. Code and processing scripts are available at https://github.com/meisamjam/GroupAffect-4; the dataset is publicly archived at https://zenodo.org/records/20037847.

Embedding models are pivotal in industrial information retrieval systems like search and advertising. However, existing pretrained models often exhibit fixed architectures and embedding dimensionalities, posing significant challenges when adapting them to diverse deployment scenarios with varying business-driven constraints. A common practice involves fine-tuning with partial parameter initialization from larger pretrained models for resource-constrained tasks. This method is often suboptimal as the misalignment between pretraining and downstream usage prevents full realization of pretraining benefits. To address this limitation, we introduce m3BERT: a Modern, Multi-lingual, Matryoshka Bidirectional Encoder, which features a novel pretraining strategy that jointly optimizes representations across both transformer layers and multiple embedding dimensions. This enables a single model to be tailored to varied resource and accuracy targets while maintaining consistency with pretraining. Incorporating recent architectural improvements, m3BERT uses a three-stage pretraining: monolingual pretraining, multilingual adaptation to serve diverse user bases, and crucial continual pretraining on a massive web domain corpus to enhance utility in commercial retrieval. m3BERT significantly outperforms state-of-the-art embedding models in Bing-Click, a large-scale industrial retrieval dataset, showcasing its practical versatility as an efficient foundation for resource-aware industrial retrieval systems. Further experiments on public datasets also confirm the general effectiveness of our multigranular Matryoshka pretraining strategy.

Reinforcement learning for legged locomotion has matured into a stack of multi-component reward functions and physics-engine benchmarks whose morphologies are uniformly derived from real commercial hardware. Game NPCs, however, are bound by stylistic constraints absent from sim-to-real robotics and routinely take the form of creatures with no real-robot counterpart. We introduce ARC-RL, a suite of four MuJoCo continuous-control environments featuring robotic morphologies inspired by the bestiary of ARC Raiders: the 18-DoF tall hexapod Queen, the 12-DoF armoured hexapod Bastion, the 18-DoF compact hexapod Tick, and the 12-DoF quadruped Leaper. All four robots share a unified observation template, action convention, simulation cadence, and a single closed-form multi-component reward function whose only per-morphology variation lives in a small set of weights and parameters. The reward fuses a velocity-tracking tent, a healthy survive bonus, a phase-locked gait-compliance bonus/cost pair, action regularisers, three safety penalties, and a posture anchor; no motion-capture data enters the reward at any point. We additionally provide hand-crafted Central Pattern Generator demonstrators per morphology, which serve both as fixed expert references and as sources of prior data for offline-to-online training. On this playground, we conduct a controlled empirical study comparing standard online algorithms (SAC, SPEQ, SOPE-EO) and methods augmented with prior data (SACfD, SPEQ-O2O, SOPE), and characterise how each paradigm copes with the playground's morphological diversity and animation-style stylistic constraints. Source code is available at https://github.com/CarloRomeo427/ARC_RL.git.

Adapting small instruction-tuned models to specialized domains often relies on supervised fine-tuning (SFT) on curated instruction-response examples, which is expensive to collect at scale. Synthetic training examples generated by a teacher LLM from a domain corpus can reduce this cost, but existing pipelines can produce homogenized outputs and do not consistently capture cross-passage or cross-document dependencies. We introduce EmbGen, a synthetic data generation pipeline that decomposes a corpus into entity-description pairs, reassembles them using semantic structure inferred from embedding similarity, and then generates question-answer (QA) pairs via proximity, intra-cluster, and inter-cluster sampling with cluster-specialized system prompts. We evaluate EmbGen against EntiGraph, InstructLab and Knowledge-Instruct on three datasets of varied semantic heterogeneity, under fixed token budgets (5 and 20 million tokens). We use lexical overlap metrics, an LLM-as-a-judge rubric, and Binary Accuracy, a composed metric combining Factual Accuracy and Completeness for evaluation. EmbGen improves Binary Accuracy on the most heterogeneous dataset by 12.5% at 5M and 88.9% at 20M tokens budget, relative to the strongest baseline, while remaining competitive across other datasets with lower heterogeneity.

In retrieval-augmented coding, failures often begin when the relevant file is absent from the retrieved context. Under frozen generic tokenization, where a BM25 index has been built by a search system whose analyzer the practitioner does not control, this failure is routine: BM25's logarithmic RSJ-odds IDF under-separates the identifier tail that distinguishes one function from another. We replace the outer logarithm of the Robertson-Spärck-Jones odds with a q-logarithm. At q=1 the transform recovers BM25 exactly by L'Hôpital's rule, and for q<1 it is a Box-Cox transform of the RSJ odds with lambda = 1-q. On CoIR CodeSearchNet Go (182K documents), oracle-tuned NDCG@10 rises from 0.2575 to 0.4874 (absolute +0.2299; +89.3% relative; zero sign reversals in 10,000 paired-bootstrap resamples, reported as p <= 10^-4). The effect is graded across code languages and is near-zero on BEIR text. A one-parameter closed form estimates a corpus-level q from hapax density and stays near q=1 on corpora where BM25 is already optimal. The index-time cost is a single pass over the sparse score matrix and query latency is unchanged. A tokenizer ablation shows that identifier-aware tokenization largely removes the incremental gain from q-IDF.

Machine Translation (MT) for Ancient Greek (AG) to Modern Greek (MG) is a low-resource task, constrained by the lack of large-scale, high-quality parallel data. We address this gap by introducing the AG-MG Parallel Corpus, a new resource containing 132,481 sentence-aligned pairs derived from literary, historical, and biblical texts. We present a novel corpus creation pipeline that combines web-scraped, excerpt-level data with a multi-stage sentence-level alignment, and refinement process. Our method uses VecAlign with LaBSE embeddings, which we first fine-tune on a manually-aligned AG-MG subset, followed by an LLM-based error/misalignment correction phase using Gemini 2.5 Flash to ensure high alignment quality. Furthermore, we provide the first comprehensive benchmark of modern MT models on this task, evaluating three fine-tuning strategies across NMT models (NLLB, M2M100) and a Greek LLM (Llama-Krikri-8B). Our experiments show that fine-tuning yields significant improvements over base models, increasing performance by up to +10.3 BLEU points. Specifically, full-parameter fine-tuning of Llama-Krikri-8B achieves the highest overall performance with a BLEU score of 13.16, while the QLoRA-adapted M2M100-1.2B model demonstrates the largest relative gains and highly competitive results. Our dataset and models represent a significant contribution to Greek NLP.

Large-scale news corpora support a wide range of research in Computational Social Science and NLP, yet access remains constrained: commercial archives impose prohibitive costs and licensing restrictions, while open alternatives like Common Crawl's CC-News require terabyte-scale storage and computationally intensive processing. We present Infini-News, a retrieval toolkit and index for the entire CC-News archive from August 2016 to the latest available snapshot. Our contributions are threefold. First, we extract, clean the text, and parse the structured metadata of over 1.35B articles. Second, we enrich the corpus with language detection using three frontier language classifiers (GlotLID, lingua, and CommonLingua), and with multi-source geographic attribution that resolves a country of origin for 83.4% of articles across 222 countries. Third, we construct Infini-gram indexes: suffix-array structures that let researchers search the full archive for arbitrary text patterns in sub-second time. Together, these resources lower the barrier to longitudinal, cross-national media research.

Somali is a Cushitic language of the Horn of Africa with ~25 million speakers, yet no documented dedicated Somali pretraining corpus with a companion tokenizer and language-identification benchmark has been publicly released. Existing Somali text appears either inside multilingual distributions (HPLT v2, CC100, MADLAD-400, OSCAR, mC4) or in small, undocumented Somali-only uploads on Hugging Face. We introduce SomaliWeb v1, a quality-filtered Somali corpus of 819,322 documents (~303M tokens) built from three upstream sources (HPLT v2, CC100, Somali Wikipedia) through a six-stage reproducible pipeline. We release (i) the corpus, (ii) a matched BPE-16K tokenizer, and (iii) the first public side-by-side Somali benchmark of three production language identifiers. Our measurements reveal concrete quality defects in existing distributions: HPLT v2's "cleaned" Somali release retains 17.3% byte-exact duplicates, 56.1% of its documents contain fixable mojibake, and 10.7% of its byte-unique documents are near-duplicates at Jaccard tau=0.80. Our BPE-16K tokenizer emits 40.2% fewer tokens than GPT-4's cl100k_base on FLORES-200 Somali devtest as a tokenizer-level measurement; downstream language-model perplexity comparisons are deferred to a follow-up release.

Multimodal large language models are increasingly used as agent backbones that understand multimodal inputs, plan retrieval actions, invoke external tools, and reason over retrieved information. Yet existing benchmarks rarely evaluate this ability in short-video applications, where a paused frame is often visually ambiguous and answering requires vertical, long-tail, and fast-evolving domain knowledge. We introduce SVFSearch, the first open benchmark for short-video frame search in the Chinese gaming domain. SVFSearch contains 5,000 four-choice test examples and 4,198 auxiliary training examples, each centered on a paused game scene from a real short-video clip. To support fair and reproducible evaluation, SVFSearch provides a frozen offline retrieval environment with a game-domain text corpus, a topic-linked image gallery, and text, image, and multimodal retrieval interfaces, avoiding reliance on uncontrolled web search APIs. We evaluate representative paradigms ranging from direct QA and RAG workflow to Plan-Act-Replan agents and learned search models. Results reveal a large gap between model-only answering, practical agentic search, and oracle knowledge: the best open-source direct-QA model reaches 66.4%, the best practical agent achieves 79.1%, and oracle knowledge reaches 95.4%. Further analysis exposes bottlenecks in visual grounding, retrieval quality, evidence-grounded reasoning, and tool-use behavior, including over-search, answer-only shortcuts, and retrieval-induced misleading.

While modern Automatic Speech Recognition (ASR) systems achieve high accuracy on benchmark corpora, their performance often degrades when there is real-world variability. This work focuses on variability arising due to accented, spontaneous, and domain-specific speech. In particular, we introduce PAper REading DAtaset (PAREDA), a first-of-its-kind multi-accent speech dataset consisting of discussions on academic Natural Language Processing (NLP) papers between speakers with Australian, Indian-English, and Chinese English accents. Each session elicits a spontaneous monologue (a summary of a paper's abstract) and a non-monologue (a question-and-answer session between participants), resulting in a corpus rich with technical jargon and conversational phenomena. We evaluate the performance of SOTA ASR models on PAREDA, analysing the impact of accent mixing and increased speech rate. Our results show that, in the zero-shot setting, models perform worse, confirming the dataset's challenging nature. However, fine-tuning on PAREDA significantly reduces the Word Error Rate (WER), demonstrating that our dataset captures linguistic characteristics often missing from existing corpora. PAREDA serves as a valuable new resource for building and evaluating more robust and inclusive ASR systems for specialised, real-world applications.

LLM pretraining is shifting from a compute-bound to a data-bound regime, where available human (organic) text falls far short of scaling demands. However, reaching the data-bound regime does not mean the model has fully utilized its organic corpus. In this paper, we introduce SynPro, a synthetic data generation framework that helps LLMs more thoroughly learn from limited organic data. SynPro applies two operations, rephrasing and reformat, that present the same organic source in diverse forms to facilitate deeper learning without introducing external information. Both generators are optimized via reinforcement learning with quality, faithfulness, and data influence rewards, and are continuously updated as pretraining plateaus to target content the model has yet to absorb. We pretrain 400M and 1.1B models with 10% of their Chinchilla-optimal tokens (0.8B and 2.2B) from DCLM-Baseline, reflecting a realistic data-bound regime in frontier pretraining. Our results reveal that organic data is significantly underutilized by standard repetition: SynPro unlocks 3.7-5.2x the effective tokens of repetition, even surpassing the non-data-bound oracle that trains on equivalent unique data at the 1.1B scale. Analyses confirm that faithful, model-aware synthesis sustains data-bound scaling without causing distribution collapse. We open-source our code at https://github.com/cxcscmu/SynPro.

Large language models pretrained on general-domain corpora often exhibit tokenization inefficiencies when applied to specialized domains. Although continual pretraining for domain adaptation partially alleviate performance degradation, it does not resolve the fundamental vocabulary mismatch. To address this gap, we introduce a targeted parameter-efficient domain adaptation approach that combines vocabulary adaptation with pretraining for LLM-based text summarization. Our unified framework augments pretrained tokenizers with domain-specific tokens while selectively replacing under-trained and unreachable tokens to limit parameter growth. We evaluate our approach on Llama-3.1-8B and Qwen2.5-7B across legal and medical summarization tasks on a challenge-oriented evaluation protocol focused on expert-driven text and summaries which typically has higher concentration of over-fragmented Out-of-Vocabulary (OOV) words. The vocabulary adaptation algorithm enhances the overall quality of the summarization model by improving semantic similarity between the generated summaries and their references. In addition, the adapted model produces summaries that incorporate more appropriate novel and domain-specific words, leading to improved coherence, relevance, and faithfulness. We further observe that our proposed approach significantly reduce training time by $35-55\%$ over continual pretraining and reduce parameter counts up to $37\%$ w.r.t expansion-only methods. We make the codebase publicly available at https://github.com/gb-kgp/VocabReplace-Then-Expand.

India's Right to Information Act, 2005 gives every citizen the right to demand information from public authorities, yet in practice most people cannot make sense of the dense administrative language used in Central Information Commission (CIC) decisions, let alone predict whether an appeal is worth filing. This paper introduces RTI-Bench, a structured dataset of CIC decisions with outcome labels, exemption citations, IRAC-style reasoning components, and procedural timelines. To the best of our knowledge it is the first publicly released structured dataset for Indian RTI administrative decisions. The dataset draws from two sources: 1,218 cases from a publicly available instruction-response corpus (with structured fields added through rule-based extraction), and 298 CIC decision PDFs collected directly from the Commission portal, spanning five commissioners and three document format generations from 2023 to 2026. Label coverage reaches 89% on the instruction-response corpus. For the PDF subset of 239 primary decisions, coverage is 51% in this first release. A random sample of 50 labelled cases was manually reviewed, yielding a label precision of 95.3%. A zero-shot Mistral 7B baseline on 100 cases gives 57.3% accuracy and 37.0% macro-F1 on outcome prediction, well above the majority-class baseline of 14.3% macro-F1. RTI-Bench is available at https://huggingface.co/datasets/joyboseroy/rti-bench

We present vega-mir, an open-source Python library that bundles nine information-theoretic and statistical metrics for the analysis of symbolic music corpora behind a small, tested, citable API, and demonstrates two of them at corpus scale in case studies not addressed by the upstream Cygnus paper. Of the nine metrics, three (Shannon entropy, Kullback-Leibler divergence, Zipfian fits) were deployed in the companion Cygnus arXiv preprint; two (network analysis on chord-transition graphs and spectral analysis of rubato curves) are deployed in full case studies here; the four remaining (multi-dimensional Gini, chi-squared stationarity, Higuchi fractal dimension, interval distribution) are validated against analytic anchors and exercised as sanity checks on a bundled 8-composer dataset. The two case studies yield two main observations. First, on the fourteen MAESTRO composers with N >= 10 pieces, the PageRank value of the gravity-centre node correlates with the marginal Kullback-Leibler distance at rho = 0.61 (Spearman, composer-level jackknife N = 14); the categorical gravity-centre identity takes five distinct values across the corpus but is not itself correlated with marginal KL (rho = 0.13, p = 0.21). Second, on the 247-piece Bach multi-master corpus (Schiff, Gould, Richter), Gould holds the highest periodicity ratio of the three performers, not the lowest, inverting the cliché that low scalar rubato reads as "metronomic": Gould's rubato is small in amplitude but structured in time, with a median dominant period of 66 beats against Schiff's 102 and Richter's 104.

This paper introduces a dataset of interlinked multimodal political communications from the Russian government, addressing persistent deficiencies in the availability of social text- and image-based data for authoritarian politics contexts. The dataset comprises two large corpora of official speeches delivered by senior actors within the Kremlin and the Russian Ministry of Foreign Affairs over multiple decades. For each speech, we provide Russian- and English-language texts, associated images and captions where available, and harmonized metadata including (e.g.) dates, speakers, (geo)locations, and official government content tags. Unique identifiers link images to speeches and align Russian and English versions of the same communication texts. We further augment these linked datasets with validated topical annotations for both speech texts and speech images, which are generated via transformer-based multimodal topic modeling and refined by a Russian politics expert. The resulting data resources support multimodal, multilingual, temporal, and/or spatial analyses of (authoritarian) political communication and offer a valuable testbed for social science research and large language model (LLM) applications in political domains.

Conversational nurse-patient transcripts contain actionable observations, but converting these transcripts into structured representations at scale remains challenging. Documentation burden is substantial, with prior studies showing clinicians spend large portions of their workday on documentation and related desk work rather than direct patient care. MEDIQA-SYNUR focuses on observation extraction from conversational nurse-patient transcripts, requiring systems to normalize these narratives into a predefined schema with value-type constraints. We propose a modular retrieval-augmented generation (RAG) pipeline that uses the training set as an exemplar corpus, combines schema-constrained prompting (full schema vs. pruned candidate schema), deterministic schema-based postprocessing, and a second-pass audit, with two LLM backbones: Llama-4-Scout-17B-16E-Instruct and GPT-5.2 with corresponding embedding models for RAG. Our best configuration uses GPT-5.2 with full schema, RAG, and a second-pass auditing, achieving 80.36% F1 score. Overall, our results show that RAG consistently improves performance, while the optimal degree of schema constraint depends on the model, and second-pass auditing yields modest additional gains by correcting residual schema-adherence errors.

Causal inference, estimating causal effects from observational data, is a fundamental tool in many disciplines. Of particular importance across a variety of domains is the continuous treatment setting, where the variable of intervention has a continuous range. This setting is far less explored and represents a substantial shift from the binary treatment setting, with models needing to represent effects across a continuum of treatment values. In this paper, we present the first causal foundation model for the continuous treatment setting. Our model meta-learns the ability to predict causal effects across a wide variety of unseen tasks without additional training or fine-tuning. First, we design a novel prior over data-generating processes with continuous treatment variables in order to generate a rich causal training corpus. We then train a transformer to reconstruct individual treatment-response curves given only observational data, leveraging in-context learning to amortize expensive Bayesian posterior inference. Our model achieves state-of-the-art performance on individual treatment-response curve reconstruction tasks compared to causal models which are trained specifically for those tasks. Inference code (including trained model weights) can be found at https://github.com/layer6ai-labs/CCPFN-inference .

Elections represent a crucial milestone in a nation's ongoing development. To better understand the political rhetoric from various movements, ranging from left to right, we propose a transformer-based model capable of projecting the political orientation of a text on a continuous left-to-right spectrum, represented by a normalized scalar d between -1 and 1. This approach enables analysts to focus on specific segments of the political landscape, such as conservatives, while excluding liberal and far-right movements. Such a task can only be achieved with multiclass classifiers, provided that the desired orientation is incorporated within one of their predefined classes. To determine the most suitable foundation model among 13 candidate transformers for this task, we constructed four distinct corpora. One corpus comprised annotated plenary notes from the German Bundestag, while another was based on an official online decision-making tool, Wahl-O-Mat. The third corpus consisted of articles from 33 newspapers, each identified by its political orientation, and the fourth included 535,200 tweets from 597 members of the 20th and 21st German Bundestag. To mitigate overfitting, we used two distinct corpora for training and two for testing, respectively. For in-domain performance, DeBERTa-large achieved the highest F1 score F1=0.844 as well as for the X (Twitter) out-of-domain test ACC=0.864. Regarding the newspaper out-of-domain test, Gemma2-2B excelled (MAE = 0.172). This study demonstrates that transformer models can recognize political framing in German news at the level of public opinion polls. Our findings suggest that both the model architecture and the availability of domain-specific training data can be as influential as model size for estimating political bias. We discuss methodological limitations and outline directions for improving the robustness of bias measurement.

We audit the multimodal-physics evaluation pipeline end-to-end and document three undetected construction practices that distort how the field measures vision-language reasoning: train-eval contamination, translation drift, and MCQ saturation. (1) Public training pools (UGPhysics-Train, SciInstruct, MMK12) pass single-stage 5-gram-Jaccard audits with zero hits across all six public physics evals; a three-stage audit (Jaccard -> mxbai-embed-large cosine -> Haiku-4.5 LLM-judge) surfaces 134 near-duplicates and 4,846 paraphrase candidates in SciInstruct alone. (2) A 17-pp Sonnet 4.5 delta on 59 paired Estonian-English olympiad problems (30.5% vs. 13.6%; sign test p=0.011, McNemar p=0.021, paired bootstrap 95% CI [+5.1, +28.9] pp). (3) A 46-pp format-and-novelty gradient on identical Sonnet weights between MCQ (79.7% on PhyX) and open-ended olympiad evaluation (33.4% on PhysOlym-A). We release four artifacts addressing these gaps: PhysCorp-A (6,432-record three-stage-audited multimodal corpus), PhysR1Corp (2,268-record closed-form RL pool), PhysOlym-A (500-problem, 99.8% novel-source held-out olympiad eval with native difficulty labels and an EN/ET bilingual subset), and Physics-R1, a reference GSPO+DAPO recipe cold-started from Qwen3-VL-8B-Thinking. Across 3 seeds, Physics-R1 lifts the audited corpus over the 8B base by +18.3 pp on PhysOlym-A liberal (8.0 -> 26.3 +/- 1.7; 7.1 pp behind Sonnet 4.5), +15.7 pp on PhysReason (23.9 -> 39.6 +/- 6.4; ahead of Qwen3-VL-32B and Gemini 2.5 Pro), +6.9 pp on OlympiadBench-Physics (46.2 +/- 1.5), and +4.1 pp on PhyX MCQ (77.8 +/- 0.3).

Bioacoustic recognition requires fine-grained acoustic understanding to distinguish similar-sounding species. However, many large-scale data repositories such as iNaturalist are weakly annotated, often with only a single positive species label per recording, making supervised learning particularly challenging. Inspired by advances in computer vision, recent approaches have shifted toward self-supervised learning to capture the underlying structure of audio without relying on exhaustive annotations. In particular, masked autoencoders (MAE) have shown strong transferability on massive audio corpora, yet their effectiveness in more modest bioacoustic settings remains underexplored. In this work, we conduct a systematic study of MAE pretraining for species classification on iNatSounds, analyzing the impacts of pretraining data scale, domain specificity, data curation, and transfer strategies. Consistent with prior work, we find that models pretrained on diverse general audio data achieve the best transfer performance on iNatSounds. Contrary to observations from large-scale audio benchmarks, we find that (1) additional masked reconstruction pretraining on domain-specific data provides limited benefits and may even degrade performance relative to off-the-shelf models, and (2) selective data filtering offers a negligible advantage when the overall data scale is limited. Our results indicate that, in moderate-sized fine-grained bioacoustic settings, pretraining scale dominates objective design. These findings further clarify when MAE-based pretraining is effective and provide practical guidance for model selection under limited supervision.

High-quality training datasets are essential for the performance of neural networks. However, the audio domain still lacks a large-scale, strongly-labeled, and single-source sound event dataset. The FSD50K dataset, despite being relatively large and open, contains a considerable fraction of multi-source samples where background interference or overlapping events could limit the usefulness of the data. To address this challenge, we introduce a data curation framework designed for large-scale open audio corpora. Our approach leverages a generative diffusion model to synthesize clean single-class events to construct controlled noisy mixtures for supervision. We subsequently employ a pre-trained audio encoder coupled with a discriminative classifier to automatically identify and filter out multi-source samples. Experiments show that our framework achieves strong performance on a human expert-curated test set. Finally, we release FSD50K-Solo, a model-curated subset of FSD50K containing single-source audio samples identified by our method. Beyond FSD50K, our method establishes a scalable paradigm for curating open source audio corpora.

This paper presents a multi-stage framework for detecting reclaimed slurs in multilingual social media discourse. It addresses the challenge of identifying reclamatory versus non-reclamatory usage of LGBTQ+-related slurs across English, Spanish, and Italian tweets. The framework handles three intertwined methodological challenges like data scarcity, class imbalance, and cross-linguistic variation in sentiment expression. It integrates data-driven model selection via cross-validation, semantic-preserving augmentation through back-translation, inductive transfer learning with dynamic epoch-level undersampling, and domain-specific knowledge injection via masked language modeling. Eight multilingual embedding models were evaluated systematically, with XLM-RoBERTa selected as the foundation model based on macro-averaged F1 score. Data augmentation via GPT-4o-mini back-translation to alternate languages effectively tripled the training corpus while preserving semantic content and class distribution ratios. The framework produces four final runs for the evaluation purposes where RUN 1 is inductive transfer learning with augmentation and undersampling, RUN 2 with masked language modeling pre-training, RUN 3 and RUN 4 are previous predictions refined via language-specific decision thresholds optimized via ROC analysis. Language-specific threshold refinement reveals that optimal decision boundaries vary significantly across languages. This reflects distributional differences in model confidence scores and linguistic variation in reclamatory language usage. The threshold-based optimization yields 2-5% absolute F1 improvement without requiring model retraining. The methodology is fully reproducible, with all code and experimental setup available at https://github.com/rbg-research/MultiPRIDE-Evalita-2026.

In this paper, we examine linguistic puzzles used in high school linguistics competitions, focusing on two common formats: Rosetta Stone and Match-Up. We propose a systematic procedure for converting existing Rosetta Stone puzzles into corresponding Match-Up counterparts. Because linguistic puzzle creation is complex and time-consuming, our method provides an efficient way to accelerate the generation of new puzzles. We evaluate the resulting Rosetta Stone-Match-Up pairs with both human participants and large language models (LLMs). Our results show that both expert human solvers and LLMs display an all-or-nothing pattern on Match-Up puzzles, either solving them completely or failing entirely. This work contributes a new dataset of paired puzzles and provides a detailed evaluation of puzzle difficulty across formats, offering insights into both human and machine linguistic reasoning.

Molecular structure elucidation from tandem mass spectra (MS/MS) remains challenging, particularly for de novo generation beyond database coverage. A common approach decomposes the task into spectrum-to-fingerprint prediction followed by fingerprint-to-structure decoding, enabling the use of large-scale molecular corpora. However, at deployment, the decoder relies on predicted rather than oracle fingerprints, introducing structured errors that propagate into generation. This results in a fundamental condition mismatch, where models trained on clean inputs must operate under noisy, biased predictions, especially for long-tail substructures. We present CoRe-Gen that explicitly addresses this gap. CoRe-Gen improves the intermediate condition via synthetic-spectrum pretraining of the encoder, matches deployment-time noise through frequency-aware fingerprint corruption during decoder training, and mitigates residual errors using structure-aware autoregressive decoding with compositional SELFIES representations, auxiliary structural supervision, and lightweight chemical constraints. Experiments on standard benchmarks show that CoRe-Gen establishes a new state of the art on NPLIB1, achieving 19.54% Top-1 and 29.92% Top-10 exact-match accuracy, while remaining competitive on the more challenging MassSpecGym benchmark. Importantly, CoRe-Gen preserves the efficiency advantages of autoregressive decoding, providing a practical and scalable solution for robust spectrum-to-structure generation under realistic conditions.

As language models scale, the amount of data they require grows -- yet many target data sources, such as low-resource languages or specialized domains, are inherently limited in size. A common strategy is to mix this scarce but valuable target data with abundant generic data, which presents a fundamental trade-off: too little target data in the mixture underexposes the model to the target domain, while too much target data repeats the same examples excessively, yielding diminishing returns and eventual overfitting. We study this trade-off across more than 2,000 language-model training runs spanning multiple model and target dataset sizes, as well as several data types, including multilingual, domain-specific, and quality-filtered mixtures. Across all settings, we find that repetition is a central driver of target-domain performance, and that mixture training tolerates much higher repetition than single-source training: scarce target corpora can be reused 15-20 times, with the optimal number of repetitions depending on the target data size, compute budget, and model scale. Next, we introduce a repetition-aware mixture scaling law that accounts for the decreasing value of repeated target tokens and the regularizing role of generic data. Optimizing the scaling law provides a principled way to compute effective mixture configurations, yielding practical mixture recommendations for pretraining under data constraints.