Bert
-
Metric-Aligned Group Relative Policy Optimization for Explainable Deepfake Detection
2026-11-10 18:08:35Luka Dragar, Peter Peer, Vitomir \v{S}truc, Borut BatageljAbstract
Deepfake detectors are commonly evaluated on real/fake classification alone, but real-world applications require explanations of the visual evidence behind each prediction. Joint vision-language models struggle with authenticity assessment, producing "real" narratives for manipulated images. To address this, we propose a decoupled predict-then-explain pipeline as our submission to the ACM Multimedia 2026 Explainable Deepfake Detection Challenge, with methods trained and evaluated on the XPlainVerse benchmark. A vision-foundation-model ensemble first classifies the image, then a Qwen3-VL-8B model generates a decision-conditioned explanation of the visual evidence. For fake images, a second Qwen3-VL-8B compresses that text into a short non-expert summary without using the image. We further apply metric-aligned Group Relative Policy Optimization (GRPO) to optimize both explainers against the challenge criteria. Under matched conditions, GRPO raises complex BERTScore-F1 from 0.686 to 0.712 and also improves grounding (entity and evidence F1); our submission ranked first on complex BERTScore-F1 and seventh overall.
Abstract (translated)
URL
https://arxiv.org/abs/2611.14436
PDF
-
An Energy-Based Perspective on Machine Learning
2026-11-10 17:37:22Bo Dai, Yilun DuAbstract
Energy-based models (EBMs) define a distribution through a scalar energy, $p_\theta(x)\propto\exp(-E_\theta(x))$, and are usually presented as one model family among many. This tutorial has a broader goal: beyond introducing EBMs, it uses them as a single lens that connects the different aspects of machine learning. Any EBM is specified by three choices, the energy, which determines what the model represents; the inference procedure, which samples from or optimizes the energy; and the learning method, which fits the energy to data. We first develop these components, including sampling, variational inference, and learning by maximum likelihood, noise contrastive estimation, pseudolikelihood, score matching, and adversarial training, and show that the learning methods share one contrastive structure. We then show that discriminative models, generative models such as GANs, normalizing flows, diffusion and autoregressive models, representation learning methods such as contrastive learning, word2vec, BERT, denoising auto-encoders, and diffusion representations, reinforcement learning with EBM policies and world models, inverse reinforcement learning, and planning, as well as inference-time computation with verifiers, are all EBMs that differ in these three choices, mostly in how they handle the partition function and sampling. This view explains existing methods, organizes them into a common framework, and suggests new algorithms by replacing one component while keeping the others, with composition of energies as a principled way to combine learned models at inference time.
Abstract (translated)
URL
https://arxiv.org/abs/2611.14383
PDF
-
StaRSE: Compact Russian Sentence Embeddings with a Sign-Coded Static Encoder
2026-11-10 15:47:46Boris Malashenko, Ivan Jarsky, Egor Soldatov, Vladislav Luzhnov, Sviatoslav Stumpf, Vladislav Shmatkov, Viacheslav Shalamov, Andrey Stankevich, Valeria EfimovaAbstract
Transformer sentence encoders provide strong retrieval quality, but their inference cost becomes a bottleneck when embeddings must be produced on CPUs or refreshed frequently. We introduce StaRSE, a compact Russian static sentence encoder that removes Transformer inference at deployment time: token directions are stored as sign bits, token norms are stored separately, and sentences are encoded by mean pooling over norm-scaled sign directions. StaRSE is initialized from ruBERT-base, fine-tuned on Russian adjacent-sentence and task-shaped pairs, and post-trained with a contrastive stage that directly optimizes the sign-projected token table. On RuMTEB, StaRSE reaches a task mean of $51.16$, $7.98$ points above the strongest included static baseline. The packaged model occupies $11.3$ MiB and has a mean CPU latency of $0.08$ ms per sentence, compared with an estimated $55.7$ MiB 16-bit parameter footprint and $1.43$ ms mean CPU latency for the closest-quality contextual encoder. Ablations examine adjacent-sentence supervision, token norms, and direct sign-projected post-training.
Abstract (translated)
URL
https://arxiv.org/abs/2611.14236
PDF
-
MISO-BR: An Open Corpus for Misogyny Detection in Brazilian Portuguese
2026-11-10 13:03:50F\'abio A. S. Passos, P\'ericles B. C. Miranda, Ana Paula CavalcantiAbstract
Misogyny is among the most prevalent forms of online hate speech directed at Brazilian users, yet open and auditable resources dedicated to it in Brazilian Portuguese remain scarce. We introduce MISO-BR, an open, documented and versioned corpus of 896 balanced social media texts annotated for binary misogyny, released with its construction protocol and reproducible baselines. The corpus combines lexicon- and target-guided scraping from X with items imported from OLID-BR, a large language model triage stage whose prompt and decoding parameters are released, and manual annotation by three annotators following a written guide. Inter-annotator reliability reaches a Fleiss' kappa of 0.788 over 816 items. Under a single fixed protocol we compare four classical classifiers (Multinomial and Bernoulli Naive Bayes, a linear SVM and a random forest) with fine-tuned BERTimbau Base and Large. The classical models cluster between 0.651 and 0.676 macro-F1 and are precision-oriented but recall-poor. The transformers reach 0.833 macro-F1 and accuracy (Wilson 95% CI for accuracy [0.77, 0.88]), a gap of roughly 16 points, while Base and Large are indistinguishable on the test set despite a 2.6x difference in training cost. We argue that the value of the resource lies less in its headline scores than in the auditability of its pipeline, and we state three limits that users must reckon with: the artificial class balance, whose effect on precision under realistic prevalence we quantify; the association between provenance and label, which the released source field makes measurable; and the selection bias induced by lexicon-guided sampling and automatic triage. The corpus, annotation guide, models and experimental scripts are publicly available.
Abstract (translated)
URL
https://arxiv.org/abs/2611.14036
PDF
-
Softness is something you also need! Why compliance helps control: information metrics of a passive walker with compliant ankle
2026-11-10 09:17:17Fabio Bonsignorio, Elisabetta Barletta, Sorin Dragomir, Francesco Esposito, Enrica ZereikAbstract
A great gap in adaptivity, flexibility, and dexterity still exists between intelligent robots, whether they are made of rigid or (partially) compliant components, and their naturally evolved animal and vegetal counterparts. One reason could be of `paradigm' nature. On the one hand, most robots (both rigid and soft) are still designed by following a `Cartesian' approach, with top-down control and a clear mind-body separation. Designing artificial agents on the basis of the principles that we observe in biology is not an easy task. The principles of organization of animals and plants rely on morphological computation and emergence of fit behaviors from loosely coupled networks of embodied agents. Modeling and control require radically new approaches. In this paper we briefly recall and explain some seminal mathematical results linking information metrics and dynamics of the agents, already published years ago; we integrate them with some recent methods developed by us and by the community, we highlight some brief remarks about their relevance and we hint to our current and near future work towards finding concrete and efficient ways for the modeling and control of completely or partially compliant artificial agents and for a better understanding of physical intelligence in nature. We present a technical note summarizing phase-space volume calculations for inverted-pendulum and passive-dynamic walking models, following our proposed framework. Starting from the McGeer compass gait, we introduce knee-ankle compliance, foot-rocker effects, and a Ruina-style compass gait with distributed leg inertia. Standard heel-strike maps are derived and used in the estimation of viable phase-space volumes $W(\gamma)$, based on a state-dependent Monte-Carlo and Reproducing Kernel Hilbert Spaces methods. We show how the methods based on RKHS allow deeper and more subtle analysis.
Abstract (translated)
URL
https://arxiv.org/abs/2611.13779
PDF
-
MOAI: Module-Optimizing Architecture for non-interactive secure transformer Inference
2026-11-10 06:36:43Linru Zhang, Xiangning Wang, Jun Jie Sim, Zhicong Huang, Jiahao Zhong, Huaxiong Wang, Pu Duan, Kwok-Yan LamAbstract
Privacy concerns have been raised in Large Language Models (LLM) inference when models are deployed in Cloud Service Providers (CSP). Homomorphic encryption (HE) offers a promising solution by enabling secure inference directly over encrypted inputs. However, the high computational overhead of HE remains a major bottleneck. To address this challenge, we propose MOAI, an efficient HE-based, non-interactive framework for secure transformer inference. MOAI gains significant efficiency improvement from: (1) a novel evaluation flow that combines column and diagonal packing with consistent strategies across all layers, eliminating expensive format conversions. (2) rotation-free algorithms for Softmax and LayerNorm that significantly reduce the number of costly HE rotations, removing 2,448 HE rotations in BERT-base inference. (3) Column packing removes rotations in plaintext-ciphertext matrix multiplications and interleaved batching further reduces the rotations in ciphertext-ciphertext matrix multiplications. MOAI uses at least 1.7x fewer HE rotations compared to the state-of-the-art works across all matrix multiplications of BERT-base. As a result, we achieve a $52.8 \%$ reduction in evaluation time compared to the state-of-the-art HE-based non-interactive secure transformer inference, THOR (Moon et al., CCS 25). We then apply MOAI on the Powerformer's framework and achieve a $55.7 \%$ reduction in evaluation time compared to Powerformer (Park et al., ACL 25), which approximates Softmax and LayerNorm with simpler functions in transformer and proposes HE-based non-interactive transformer inference. We report an amortized time of 2.36 minutes per input on a single GPU environment. We show the extendibility by applying MOAI in LLaMA-3-8B. Our implementation is publicly available as open source.
Abstract (translated)
URL
https://arxiv.org/abs/2611.13641
PDF
-
An Interpretable Stylometric Framework for Early Chinese Buddhist Translator Styles: A Case Study on the Attribution of Disputed Works of An Shigao
Abstract
Attributing early Chinese Buddhist translations requires evidence that scholars can trace back to the texts. We develop an interpretable stylometric framework that uses features defined or reviewed by Buddhist philologists, including function characters, sentence formulas, sentence-final characters, sentence length, and part-of-speech n-grams. We construct stylistic profiles from 31 authenticated translations by seven translators from the Eastern Han through the Western Jin, then compare eight disputed texts with An Shigao's profile. Two texts, T1557 and T0101, show the strongest stylistic similarity to the authenticated An Shigao corpus, with scores of 0.783 and 0.687, respectively. The attributions agree with those reported by a published BERT-based approach on the same set of disputed texts. The framework requires no task-specific training and allows scholars to inspect which linguistic features support each proposed attribution.
Abstract (translated)
URL
https://arxiv.org/abs/2611.13636
PDF
-
IFRS Compliance Verification: An Open Multi-LLM Benchmark and Content-Conditioned Ensemble Classifier
2026-11-10 05:12:13Aghoghomena Akasukpe, Li Yang, Xinyao Zhou, Chen Shen, Miguel Vargas MartinAbstract
Verifying corporate financial disclosures against the International Financial Reporting Standards (IFRS) is a core auditing task in which experts cross-reference disclosures with hundreds of fine-grained regulatory paragraphs: a laborious, costly, and error-prone process that motivates growing interest in large language models (LLMs) as compliance assistants. Yet single-model accuracy remains insufficient for high-stakes use, and naive ensembling is vulnerable to systematic per-model biases that can drag aggregate accuracy below that of the strongest individual annotator. This gap is addressed with a multi-stage framework, supported by a new benchmark of 19,998 IFRS regulation-disclosure pairs comprising 11,161 unanimous multi-LLM labels and 8,837 human annotations on items where the LLMs disagreed. The framework couples a three-stage domain-adapted sentence embedding pipeline (BERT masked language modeling on SEC EDGAR filings, unsupervised TSDAE and SimCSE adaptation, and supervised contrastive alignment) with a learned ensemble classifier that fuses frozen embeddings and LLM probability features from GPT-4.1, Claude Sonnet 4, and Gemini 2.5 Pro via Tree-structured Parzen Estimator (TPE) optimized gradient-boosted trees. On the disagreement set, the ensemble achieves 81.7% weighted F1, exceeding the strongest single LLM by 10.4 points and majority voting by 14.6 points, and outperforms eleven unsupervised aggregation baselines; the retrieval pipeline attains Recall@15 of 45.3% on unseen regulations. The dataset, model checkpoints, annotation platform, and classifier implementation will be publicly released.
Abstract (translated)
URL
https://arxiv.org/abs/2611.13595
PDF
-
On the Effectiveness and Limitations of Open-Weight LLMs for Low-Resource Offensive Language Detection in Danish
2026-11-06 14:23:41Elnur Adl Zarabi, Amit Kumar, Adi Singh, Sierra Nelson, Megan Stubbs-Richardson, Taoran JiAbstract
We study the effectiveness and robustness of open-weight large language models (LLMs) for Danish offensive language detection, an underexplored low-resource task. We evaluate four model families (Llama, Mistral, Gemma, and Qwen) across eleven prompting strategies spanning five design categories, cross-prompt generalization, Danish orthographic variation, and robustness under balanced and naturally imbalanced class distributions. We compare against three BERT-family baselines with distinct pre-training strategies, evaluated via both masked language model prompting and supervised fine-tuning. Even without task-specific fine-tuning, LLMs can surpass all three fine-tuned BERT baselines, and lightweight finetuning yields substantial further gains. No single model or prompt category dominates, underscoring that prompt design interacts with model pre-training in ways that cannot be predicted a priori. Semantically grounded prompts transfer more robustly, while format-constrained patterns consistently fail to generalize. ASCII transliteration of Danish-specific characters (æ, ø, å) generally degrades performance, indicating these characters encode meaningful signal. At the same time, the gap between a balanced test set (best F1 of 0.84) and the corpus's natural class skew (0.62 trained on skewed data, 0.70 trained on balanced data) -- relative drops of 26.2\% and 16.7\% -- shows that class imbalance remains a critical open challenge. Our code and configurations are available at this https URL.
Abstract (translated)
URL
https://arxiv.org/abs/2611.09506
PDF
-
Asymmetric Correspondence Learning for Lexical and Style Embeddings: Interpreting Latent Space via Acoustic Features
2026-11-06 11:04:58Michael Whealing, Robert Gaizauskas, Thomas HainAbstract
Self-supervised speech embeddings jointly encode lexical, speaker, and prosodic information, complicating feature isolation. We introduce Asymmetric Correspondence Learning (ACL), distilling HuBERT features into separate Acoustic Word Embeddings (AWEs) and Acoustic Style Embeddings (ASEs) via a split-latent correspondence bottleneck. Through probing and clustering evaluations on single-word utterances, we demonstrate that substantial variance from extrinsic sources remains encoded within each latent space, despite linear probing indicating the information has been stripped. Whilst the AWE space exhibits cohesive lexical clustering (AP$\approx$0.90, Fisher$\approx$6.28), linear probing on the ASE space achieves 97.9\% speaker accuracy across a diffuse topology (Fisher$\approx$1.01), yet preserves substantial lexical variance (Fisher$\approx$0.86) that remains linearly unrecoverable (0.17\%). To explain these findings, we apply group-level permutation feature importance over eGeMAPS acoustic features, finding that AWEs partition by vocal tract filters and ASEs by $F_0$ dynamics. These findings suggest that standard diagnostic benchmarks may be insufficient for verifying feature isolation, as substantial speech structure can be retained even when it remains entirely hidden from probes.
Abstract (translated)
URL
https://arxiv.org/abs/2611.09313
PDF
-
PhoPunct: A Large-Scale Benchmark of Transformer Architectures for Vietnamese Punctuation Restoration
2026-11-06 07:39:32Khoa D. Nguyen, Thai Bao Huynh, Binh T. NguyenAbstract
Vietnamese punctuation restoration is often evaluated on relatively small benchmarks, limiting conclusions about how architectural choices scale with data volume. We present a controlled benchmark on News and Novels corpora containing 22.1 million labeled tokens (19.9M News and 2.2M Novels). We evaluate 17 architectures formed from four pretrained Transformer backbones and optional bidirectional long short-term memory (Bi-LSTM) and conditional random field (CRF) layers, together with a PhoBERT-large-Bi-LSTM-CRF model, PhoPunct. Every model is trained for eight epochs with the same learning rate and without architecture-specific tuning. Under this fixed budget, adding Bi-LSTM and CRF layers does not reliably improve macro F1: in seven of eight matched domain-backbone comparisons, a plain Transformer classifier matches or outperforms its Bi-LSTM-CRF counterpart. PhoPunct obtains the best News result (61.35), whereas XLM-R with a CRF obtains the best Novels result (50.18), exceeding PhoPunct (48.97). These results show that the value of recurrent and structured-prediction heads depends on their optimization regime and motivate explicit separation between architectural improvements and improvements due to architecture-specific tuning. Our source code is available at: this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2611.09107
PDF
-
The Invisible Gap: Romanized Indic Generation Fails Where Automatic Metrics Succeed
2026-11-06 07:17:32Madhu Sudhan Rao Beesetty, Surbhi Sharma, Devesh Maheshwari, Lokesh MeesalaAbstract
Hundreds of millions of Indians communicate with digital systems in romanized Latin characters -- typing "aap kaise hain" rather than native script -- yet romanized Indic generation is absent from NLP benchmarks and evaluation frameworks. We demonstrate that automatic metrics (BGE-M3, BERTScore) systematically misrepresent the quality of romanized Hindi, Telugu, and Kannada generation. We evaluate four LLMs (Claude Opus 4.8, GPT-5.5, Gemini 2.5-flash, Qwen3.5-35B) on 142 OpenAssistant query-response pairs. On a 72-response diagnostic subset, native speakers rate GPT-5.5 1.43 points higher than Claude Opus 4.8 on a 12-point rubric, whereas their full-dataset BGE-M3 means differ by only 0.004. Two distinct model-specific failure modes emerge: Gemini produces script spillage (16.9% of Telugu responses leak native script), while Qwen collapses to English-only output (16.2% of Kannada responses). We introduce a failure-aware evaluation policy that assigns zero to empty outputs, script-corrupted responses, and instruction collapses -- revealing failures metrics alone cannot detect. Our results show that evaluating romanized Indic generation requires a coordinated bundle: semantic similarity, script compliance, failure coverage, and native-speaker acceptability.
Abstract (translated)
URL
https://arxiv.org/abs/2611.09085
PDF
-
Beyond Truncation? Entailment-Licensed Elision and the Limits of Semantic Preservation in LLM Decoding
2026-11-06 04:39:25Mahmoud Alyosify, Mirna Imbabi, Eman ElkhamisiAbstract
Instruction-tuned large language models (LLMs) answer at greater length than a request needs. We ask whether a training-free controller inside the decoding loop can remove that surplus while preserving content, judged against the simplest possible baseline: truncating the model's own answer to the same length. An entropy-gated logits processor removes 25.5-34.9% of tokens on MT-Bench development questions and 19.9-34.2% on a sealed test split within pre-registered BERTScore non-inferiority margins, yet we prove and confirm by audit that its output is always a prefix of the unmodified answer: the saving is truncation. A survey of 23 training-free mechanism families finds none that selects content better than truncation without oracle labels. We therefore introduce Entailment-Licensed In-Decoding Elision (ELIDE), which, at each sentence boundary, withholds a span from the reader only when a natural-language inference model judges it entailed by the text already shown, while leaving the model's generation unchanged. ELIDE beats matched-length truncation on three model families, but reading 936 of its elisions on 670 public prompts shows that 59% harm the answer, a failure that no automatic measure detects. Scoping the licence to its discourse unit reduces the harmful share to about one elision in ten on two fresh tests.
Abstract (translated)
URL
https://arxiv.org/abs/2611.08975
PDF
-
CardioLM: A multilingual suite of small language models for the cardiology domain
2026-11-05 12:41:19Manuela-Daniela Danu, Miguel Rodr\'iguez-Ortega, Alberto Becerra-Tome, Pablo Arancibia, Fernando Gallego-Donoso, Lorenzo Pratesi, Xi Bai, Andrei-Damian Dinescu, Salvador Lima-L\'opez, Andrea Guala, Saskia Haitjema, Jonas Mack, Michele Miranda, Wesam Al Nabki, Jan Rodr\'iguez-Miret, Vasile George Marica, Vladim\'ir Kincl, Kate\v{r}ina Dost\'alov\'a, Michael Andrej, Alicia Ujil, Corinne Allaart, Machteld Boonstra, Malin Overmars, Anoop Shah, Giovanna Ceroni, Jonathan Kossoff, Domenico D'Amario, Renzo Laborante, Antonio Iaconelli, Isabella Fumarulo, Jacopo Lenkowicz, Livia Lilli, Emma Weiss, Alex Scafa, Felix Lindberg, Milou Gamage, Eul\`alia Farr\'e-Maduell, S\'ebastien Brati\`eres, Martin Krallinger, Bram van EsAbstract
Background: Cardiovascular care generates large volumes of clinical free-text documentation in local languages. However, clinical natural language processing (NLP) resources remain concentrated in English, and no existing benchmark is cardiology-specific, multilingual, and broad in entity coverage. Methods: We present two contributions. The first is CardioCCC, a corpus of 508 cardiology case reports originally published in Spanish and extended to Czech, Dutch, English, Italian, Romanian, and Swedish through machine translation, annotation projection, and clinician review, with annotations covering diseases, symptoms, procedures, and medications. The second is CardioLM, comprising CardioBERTa, a set of language-specific encoders obtained by continued pretraining on cardiology and biomedical corpora, and CardioNER, a set of named entity recognition (NER) models fine-tuned on CardioCCC. Results: Cardiology-specific pretraining reduced pseudo-perplexity by up to a factor of eight, yet changed strict span-level F1 by at most 0.04, with benefits concentrated in languages whose backbones lacked biomedical coverage. The jointly trained multilingual XLM-RoBERTa-large models achieved the best strict F1 in nearly every language and entity type, exceeding the strongest monolingual system by up to 0.13 F1. A large share of residual errors concerned boundary detection: relaxed matching recovered 0.10-0.23 F1 for disease, procedure, and symptom, but only 0.01-0.07 for medication. Clinicians rated CardioNER well above a rule- and dictionary-based baseline, with the mean rating across entity types rising from 26% to 73%. Conclusions: Specialty-specific pretraining is not the decisive factor for downstream entity recognition, whereas multilingual joint training yields the strongest results in nearly all settings and a simpler deployment path for multi-country clinical information extraction.
Abstract (translated)
URL
https://arxiv.org/abs/2611.07940
PDF
-
Bridging Structure and Semantics: Knowledge Graph Verification via Triple-to-Text Alignment
2026-11-05 10:19:27Ziang CuiAbstract
Semantic accuracy is an important dimension of knowledge graph quality. Graph-based scoring methods exploit relational structure, while textual representations allow classifiers to draw on knowledge encoded by pre-trained language models (PLMs). We present GenVer, a generative verifier that scores triples using both their serialized form and generated natural language representations (NLRs). A triple-text alignment procedure selects training pairs for a sequence-to-sequence generator; a BERT-based classifier then combines features from each triple and its generated NLR. On four FactFB variants, constructed from FB15k using perturbed negative triples, GenVer obtains the highest classification accuracy among the evaluated baselines, ranging from 88.56% to 91.52%. The experiments suggest that generated representations can help triple classification, while the case analysis also exposes semantic drift and high-confidence errors.
Abstract (translated)
URL
https://arxiv.org/abs/2611.07780
PDF
-
Enabling AWS Cloud FPGA Development Using High-Level Vitis HLS
Abstract
We present a toolchain for development of ML models on AWS EC2 F2 cloud FPGA infrastructure from high-level language, implementing a complete Vitis HLS to HDK bridge to AFI workflow that has been absent from the academic literature despite eight years of AWS FPGA availability. We introduce a four-level methodology (FP32 PyTorch, bit-accurate quantized Python, HLS C-simulation, physical FPGA) to aid in localizing errors, including FPGA-specific circuit errors. Together these tools allow coding agents to effectively implement FPGA designs from user directions, testing and debugging as needed. We document the deployment workflow, timing closure strategies, and practical HLS engineering decisions that proved crucial. We validate using TinyBERT fine-tuned for address Named Entity Recognition, implementing the embedding layer and the first of four encoder layers on the FPGA hardware. Our implementation streams weights from host DRAM, yielding poor performance compared to CPU/GPU implementations; we discuss forward paths.
Abstract (translated)
URL
https://arxiv.org/abs/2611.06829
PDF
-
Specificity Bias and the Wikipedia Paradox: An xAI Analysis of AI Text Detectors under Domain Shift
2026-11-02 15:22:34Nico Morloher, Johannes Reschke, Mohamed MadyAbstract
The widespread adoption of Large Language Models (LLMs) demands reliable detection of human-authored versus synthetic text. However, current Transformer-based detectors often experience severe performance degradation during domain shifts and operate as opaque classifiers. This paper investigates the robustness and interpretability of fine-tuned Transformer models (DistilBERT, RoBERTa) by evaluating them against a custom out-of-distribution (OOD) dataset of hyper-local news synthesized via open-source LLMs (Llama 3, Mistral, Gemma). To analyze model decision boundaries, we apply a tripartite Explainable AI (xAI) framework using SHAP. Our analysis reveals three key phenomena: First, detectors often rely on a "Call-to-Action" syndrome inherent to instruction-tuned LLMs rather than generic machine text. Second, we observe a "Specificity Bias," where LLMs successfully bypass detection by hallucinating specific named entities and mimicking conversational imperfection. Finally, we illustrate the "Wikipedia Paradox," wherein formal, objective human writing is frequently misclassified as AI due to its low statistical burstiness. These findings demonstrate that purely statistical detection paradigms face systematic limitations in highly structured human domains, emphasizing the critical need for semantically-aware, interpretable detection architectures.
Abstract (translated)
URL
https://arxiv.org/abs/2611.03280
PDF
-
An Experimental Study of Hybrid Retrieval, NLI and Open-source LLMs for Scientific Claim Verification
2026-11-02 13:12:13Lakshmi Narayanan P, Narenkarthik K, Naren Kumar CAbstract
Claim verification is the task of deciding whether a source text supports a claim, contradicts it, or lacks sufficient evidence. Natural Language Inference (NLI) models can judge whether a single sentence supports or contradicts a claim, but they struggle when the evidence is spread across several sentences. Conversely, Large Language Models (LLMs) can reason across sentences, but may draw conclusions the text does not support. In this experimental study, we test whether a conservative fusion rule, labeling a claim Support or Contradict only when both models agree and defaulting to Not Enough Evidence otherwise, combines their strengths. We construct a benchmark of 2,035 claims from SciFact and CliniFact (858 Support, 152 Contradict, 1,025 Not Enough Evidence). We retrieve the top five most relevant sentences per claim using a hybrid of semantic and lexical (BM25) retrieval, which slightly outperforms either method alone. These retrieved sentences are then provided to every verification model. We then compare a DeBERTa-v3-base-mnli-fever-anli NLI baseline against Llama 3.1 8B, Qwen3 8B, and Mistral 7B. Qwen3 achieves the highest Macro-F1 (0.7937), followed by Llama 3.1 (0.7204) and Mistral (0.6388), while the NLI baseline reaches only 0.3157. Contrary to our hypothesis, fusion degrades performance, dropping Macro-F1 to 0.7392, 0.6730, and 0.5991, respectively. To understand this decline, we analyze agreements, disagreements, and common error patterns between the NLI and LLM verifiers. Our error analysis shows that NLI and LLM disagreements often result in correct LLM predictions being overturned.
Abstract (translated)
URL
https://arxiv.org/abs/2611.03093
PDF
-
Distilling Literary Sentiment from LLM Annotations into an Ordinal BERT Model
Abstract
The scarcity of human-annotated literary corpora limits the development of domain-specific sentiment analysis models. This study investigates whether LLM-generated annotations can supervise smaller models for literary sentiment analysis. We fine-tune five BERT models using ordinal regression on 10,889 literary sentences annotated by Gemma 4 31B on a ten-point sentiment scale. The models are evaluated against 1,923 human-annotated sentences from Hemingway's "The Old Man and the Sea", alongside the teacher model and conventional baselines. BERT-large achieves a Kendall's $\tau$ of 0.589, approaching the teacher's 0.604. Although a constant-neutral predictor achieves the lowest overall mean absolute error (MAE), BERT-base outperforms the teacher on non-neutral sentences, achieving an MAE of 0.726 versus 1.116. These findings demonstrate the potential of LLM-generated supervision for developing smaller literary sentiment models, while highlighting the limitations of aggregate error metrics on predominantly neutral datasets.
Abstract (translated)
URL
https://arxiv.org/abs/2611.02985
PDF
-
EffComp: Efficient Prompt Compression via Hybrid Reinforcement & Supervised Learning
2026-11-02 08:46:28Romtham Tungsununtham, Piyawat Lertvittayakumjorn, Atiwong Suchato, Peerapon VateekulAbstract
Prompt compression is aimed at reducing input prompt lengths to enable cheaper, faster, and potentially higher-quality LLM predictions. However, existing prompt compression methods are often limited by modest compression gains, a risk of hallucination, and/or high compression latency. We propose EffComp, an Efficient prompt Compression framework for RAG-based question answering (QA). EffComp employs BERT-style document reranker and sentence selector models to allow fast extractive prompt compression at the sentence level. Its extractive nature prevents hallucinations in the compressed prompts. Additionally, the training process using a hybrid reinforcement and supervised learning approach is designed to optimize the compression ratio while preserving LLM accuracy. Experiments on various QA datasets demonstrate that EffComp outperforms state-of-the-art prompt compression methods achieving the best trade-off between compression ratio and accuracy with minimal latency. This makes it practical for real-world applications.
Abstract (translated)
URL
https://arxiv.org/abs/2611.02766
PDF
