The paper critiques current optimization-based methods for balancing modalities in Multimodal Sentiment Analysis, arguing they overpromise and underdeliver. It introduces a unified evaluation framework that tests gradient- and loss-based balancing strategies, provides a theoretical diagnosis showing these methods conflate fitting speed with discriminative contribution, and proposes a research agenda for held‑out discriminative modality valuation. Experiments on CMU‑MOSI and CMU‑MOSEI demonstrate that no strategy consistently outperforms Late Concatenation, performance is highly sensitive to hyperparameters, and ratio calibration does not yield reliable gains, highlighting that loss is not utility and gradients are not importance.
By Ioanna Kaffeza, Efthymios Georgiou, Alexandros Potamianos
The paper introduces UniLID, a lightweight language identification method that uses the UnigramLM tokenization algorithm to predict a string’s language by evaluating which language’s unigram distribution best explains the text. UniLID is data‑ and compute‑efficient, allows incremental addition of new languages without retraining, and can be integrated into existing tokenization pipelines. Experiments show competitive performance against baselines such as fasttext, GlotLID‑M, and CLD3, achieving 69% accuracy with five labeled samples per language and 89% with 25, and delivering significant gains on fine‑grained dialect identification.
By Clara Meister, Ahmetcan Yavuz, Pietro Lesci, Tiago Pimentel
The paper surveys how large language models (LLMs) are being applied in legal tasks such as judgement prediction, document analysis, and drafting. It reviews the benefits of automation while highlighting legal challenges like privacy, bias, and explainability. The authors also discuss data resources for legal domain specialization and outline future research directions.
By Zhongxiang Sun
The paper introduces a lightweight evaluation method for logical reasoning in transformer-based language models, using query-key alignments within attention heads to compute a QK-score. This single forward-pass technique identifies valid versus invalid inferences and is validated across multiple reasoning benchmarks, showing robustness to distractors and deeper reasoning. Experiments span models from 1.5B to 70B parameters, demonstrating scalability.
By Eduard Tulchinskii, Anastasia Voznyuk, Laida Kushnareva, Andrei Andriiainen, Irina Piontkovskaya, Evgeny Burnaev, Serguei Barannikov
The paper introduces a multimodal in‑context learning framework that uses contrastive demonstration modeling to align large language models’ responses with the required reasoning paths. By contrasting suboptimal and better responses and incorporating a response‑conditioned retrieval mechanism, the method explicitly guides models beyond surface imitation. Experiments on various multimodal tasks, especially visual question answering, show consistent performance gains.
By Mingbo Yang, Wenqiang Wang, Zhaolu Kang, Peng Chen, Yannan Chen, Sunshang Wang, Yan Xiao
LoopVAE introduces a recurrent depth architecture that reuses a scale‑ and loop‑conditioned core across different spatial scales while keeping resolution‑changing transitions separate. The four‑block core applies 28 block operations per encoder or decoder, enabling a 29M‑parameter convolutional model to achieve 0.28 rFID and 32.54 dB PSNR on ImageNet‑256 with roughly 65% fewer parameters than comparable VAEs. Experiments with both convolutional and Transformer operators, as well as ablations on parameter sharing, demonstrate competitive image quality metrics and reveal how targeted loop interventions and truncation affect reconstruction quality and computational trade‑offs.
By Zhiying Lu
The paper investigates correctness‑gated multi‑teacher distillation, comparing a weighted arm to unfiltered distillation across eight experimental arms. While the weighted arm shows modest gains in accuracy (+0.1660) and macro‑F1 (+0.1323) and a reduction in unsafe action rate (−0.4979), it also exhibits lost label functionality, such as zero Refuted recall and over‑assignment of NotEnoughInfo. A subsequent grounding audit was inconclusive, failing to demonstrate a clear improvement in evidence grounding or overall system performance.
By Xiaofei Feng
The study investigates whether cross‑lingual clinical annotation projection can be treated as a constrained text‑generation task that preserves the original text while inserting entity tags. Using a workflow that embeds tags directly into immutable target‑language text and then validates them deterministically, the authors evaluated this approach against supervised candidate‑span projection and hybrid ML‑LLM refinement across six languages. Results show that direct LLM projection, particularly with GLM 5.2 and Gemma4:31B, achieves the highest strict F1 scores (up to 0.9201) and outperforms previous methods by 0.0564–0.1512, producing over 55,000 grounded mentions with accurate offsets.
By \'Alvaro Rey-Blanes, Francisco J. Moreno-Barea, Francisco J. Veredas
The paper introduces BodyCam-VQA, an adaptive visual question answering framework designed to improve captioning of police body‑worn camera footage. By employing structured multimodal reasoning and probe question generation, the system extracts fine‑grained visual evidence that conventional captioning models miss. Experiments with various question generation models show that this VQA‑driven approach yields more reliable, objective, and detailed records of enforcement events.
By Karish Gupta, Matthew Alex, Alex Li, Yang Wu, Yun-Wei Chu, Kashif Munir, Xiaotian Zhou, Zhengping Ji, Xiaozhong Liu
The paper introduces a framework to evaluate and diagnose the robustness of low‑resource multilingual text‑to‑speech systems when faced with complex text inputs such as numbers, dates, named entities, long sentences, code‑switched expressions, and punctuation structures. It assesses robustness across content consistency, language consistency, and generation stability, and proposes automatic metrics (character error rate, language ID accuracy, duration abnormal rate) along with a lightweight Text Risk Score (TRS) that predicts synthesis risk from interpretable text features. Experiments on Thai, Vietnamese, Swahili, and Indonesian TTS systems reveal distinct failure patterns and show that TRS correlates positively with content and duration errors, offering a low‑cost pre‑synthesis risk indicator.
By Tianlun Zuo, Ziyu Zhang, Tingzhi Mao, Zhonghua Fu, Lei Xie
The Eloquence team presents three methods for the Interspeech 2026 MLC‑SLM Task 2, a multilingual MCQA challenge covering 21 languages. They fine‑tune Voxtral‑Mini‑3B with LoRA and data augmentation, achieving 0.72 macro‑accuracy; they use multimodal in‑context learning on Voxtral‑24B to correct label bias, reaching 0.81; and they deploy a training‑free retrieval system with a voice‑anchored memory, scoring 0.68. All approaches surpass the official baseline.
By Jordi Luque, Lorenzo Concina, Marco Matassoni, Alessio Brutti, Filippo Vella
The paper introduces Narrative Consolidation, a new NLP task that aims to merge overlapping narrative documents—such as legal testimonies or historical accounts—into a single, chronologically coherent text, rather than merely compressing them. It defines the task, proposes an evaluation framework, and presents the Gospel Consolidation Language Resource, a benchmark built from the four Biblical Gospels with 169 canonical events and cross‑document alignments. Experiments show that providing an explicit temporal backbone dramatically improves performance, a simple length heuristic outperforms graph‑based methods, and temporal edges are the key discriminative signal.
By Roger A. Finger, Eduardo G. Cortes, Sandro J. Rigo, Gabriel de O. Ramos
The paper evaluates membership inference attacks (MIAs) on NLP text classifiers using the GLUE SST‑2 sentiment dataset. It compares a TF‑IDF + Logistic Regression pipeline with a fine‑tuned DistilBERT model under a loss‑threshold MIA, finding that both models leak membership signals despite high accuracy. The study also tests mitigations, showing that stronger regularization reduces leakage for Logistic Regression at a utility cost, while fine‑tuning DistilBERT for fewer epochs lowers leakage with minimal accuracy loss.
By William Novak (Minot State University), Muhammad Abusaqer (Minot State University)
HiPerViT is a compact vision-only architecture that injects an explicit second-order statistical prior into a transformer-based pipeline for texture recognition. It combines global and local image views with a compact bilinear descriptor encoded as a statistical token, and integrates this token with first-order spatial representations through Perceiver-style latent distillation. Across six texture recognition benchmarks, HiPerViT consistently outperforms strong vision-only baselines, achieving notable gains on DTD, GTOS-Mobile, and 1200Tex, and the improvements are largely independent of backbone depth or fusion topology.
By Jo\~ao Pedro C. A. de S\'a, Odemir Martinez Bruno
The paper introduces an interdisciplinary framework that uses AI and machine learning to analyze police body‑worn camera footage from the Rochester Police Department. It combines image, audio, and natural language processing—including speaker separation, transcription, and large language models—to detect and classify interaction patterns such as respect, disrespect, escalation, and de‑escalation. A custom evaluation pipeline assesses transcription quality and behavior detection accuracy, aiming to support law‑enforcement review, training, and accountability.
By Anita Srbinovska, Angela Srbinovska, Vivek Senthil, Jonathan Bateman, Adrian Martin, John McCluskey, Ernest Fokou\'e
The study examines whether financial sentiment tools that are validated against human labels also reliably predict market outcomes. Using a large corpus of securities class action messages linked to abnormal stock returns, the authors compare five sentiment instruments—VADER, Loughran‑McDonald, FinBERT, Twitter‑RoBERTa, and an LLM annotator—within a single pipeline. Results show that the alignment between human agreement and sentiment scores varies with sampling strategy and time horizon: conventional sampling favors same‑day associations, while fixed‑n panels yield similar correlations for both same‑day and one‑day‑ahead predictions, yet overall predictive rankings remain weak.
By AS Aravinthkakshan, Laven Srivastava, Harsh Nandwani
The paper introduces a multi‑signal pipeline for detecting hallucinations in large language models, combining fine‑tuned DeBERTa‑v3 classification, Monte Carlo Dropout uncertainty, and temperature‑scaled calibration. On the HaluEval benchmark it achieves high performance (F1 = 0.915, AUROC = 0.977) across QA, summarization, and dialogue, and shows that 25 % of training data yields 77 % of full‑data performance. The authors also demonstrate that applying Direct Preference Optimization to a Qwen2.5‑0.5B generator cuts hallucination rates from 85.5 % to 37.7 %, and that domain‑specific fine‑tuning (PubMedBERT on SciFact) outperforms general‑domain models for biomedical text.
By Varun Teja Chundru, Debasmita Biswas
The paper presents a method for robust multimodal sentiment analysis that handles incomplete or noisy modalities. It introduces a completeness estimation technique to measure how much sentiment-relevant information remains in partial data, guiding the reconstruction of missing semantics. A joint training strategy stabilizes multi-task learning for sentiment prediction and completeness estimation, and experiments on three benchmark datasets show improved semantic reconstruction and sentiment accuracy.
By Han-Jun Choi, Byunggill Joe, Saim Shin, Jin Yea Jang
SG-Blend introduces a per‑layer adaptive activation that interpolates between a bias‑corrected, parametric Swish variant (SSwish) and GELU, using a learnable blend coefficient, sharpness, and zero‑centering bias. The method adds only three scalars per feed‑forward block and, on BERT‑style IMDB classification, matches peak accuracy while reducing seed‑to‑seed variance by 42 %. It also achieves the lowest validation perplexity on WikiText103 and generalizes to computer vision and other domains.
By Gaurav Sarkar, Syed Affan Daimi, Jay Gala, Subarna Tripathi
The paper evaluates a multilingual ASR model (MMS‑1B‑all) on a Garrusi Kurdish dataset using a common‑reference staged normalization approach. By normalizing both reference and hypothesis, the authors show that raw Arabic‑script hypotheses yield a 111.70 % WER, which drops to 97.85 % after folding into a reduced orthography, highlighting the impact of orthographic differences on error measurement. A Southern Kurdish fine‑tuned system performs worse, and residual errors are partly due to scoring‑pipeline limitations rather than recognition failures.
By Hiwa Asadpour