Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv:2609.24560v1 Announce Type: new Abstract: Confocal Laser Endomicroscopy (CLE) provides real-time, cellular-resolution optical biopsy but has a narrow field of view, which image mosaicing can ex...
arXiv:2609.24564v1 Announce Type: new Abstract: CLIP, a foundational vision-language model, has emerged as a powerful tool for open-vocabulary semantic segmentation. While freezing CLIP's text encode...
arXiv:2609.22277v1 Announce Type: cross Abstract: Visual impairment affects over 2.2 billion people worldwide, yet conventional white canes cannot detect elevated hazards or provide semantic environm...
arXiv:2609.24576v1 Announce Type: cross Abstract: Modern Vision-Language Navigation (VLN) models rely mostly on pre-trained large Vision-Language Models (VLMs) to predict navigation actions. While th...
arXiv:2605.04515v2 Announce Type: replace Abstract: Video Large Language Models (Video-LLMs) excel in general video understanding but often base physical judgments on event expectations rather than o...
arXiv:2605.24625v2 Announce Type: replace Abstract: Ultra-low-field (ULF) MRI offers portable and accessible neuroimaging but suffers from reduced signal-to-noise ratio and limited spatial resolution...
arXiv:2606.25483v2 Announce Type: replace Abstract: Path-traced synthetic stereo is a primary training substrate for disparity networks, and the pipelines that consume it assume Monte~Carlo (MC) rend...
arXiv:2607.11295v2 Announce Type: replace Abstract: Biomedical imaging data exhibit substantial acquisition variability, where identical biological structures can appear markedly different due to dif...
arXiv:2608.08381v2 Announce Type: replace Abstract: Motivated by the IEEE 802.11bf effort to standardize advanced WLAN sensing, interest in Wi-Fi Channel State Information (CSI) for passive, device-f...
arXiv:2609.07780v2 Announce Type: replace Abstract: Automated drone surveillance has become increasingly important for public safety, critical infrastructure protection,and restricted airspace monito...
arXiv:2504.00919v3 Announce Type: replace-cross Abstract: We study the problem of estimating the spectral density of a centered stationary Gaussian time series under local differential privacy constr...
The paper investigates whether the stable GAN architecture R3GAN can improve time‑series imputation when adapted to 1‑D temporal data. Using a coarse‑to‑fine refinement framework and a frequency‑domain discriminator, the authors evaluate 14 saved configurations across three datasets and find a negative result: most configurations either show negligible improvement or degrade performance compared to baseline methods. The study highlights that the usual argument—GANs optimize distributional objectives rather than point‑wise ones—does not fully explain the lack of benefit, and it poses an open problem regarding why a learned discriminator fails to provide useful refinement gradients while diffusion denoisers succeed, offering practical guidance on when adversarial refinement may be worthwhile.
The paper introduces AFOR, a tensor‑wise adaptive optimizer for EEG decoding that replaces the fixed second‑moment decay coefficient used in Adam/AdamW with a dynamic coefficient estimated online from local gradient state. AFOR combines a Residual‑Alignment Signal Scorer (RASS) to assess gradient quality and an Adaptive Forgetting Controller (AFC) to map this score to a bounded per‑step decay coefficient, with cumulative‑product initialization correction for consistency. In cross‑subject experiments on three EEG benchmarks, AFOR outperforms standard optimizers, improving mean test accuracy by 3.00%, 2.07%, and 4.38% over Adam.
The paper compares latent representations in Selective State Space Models (SSMs) like Mamba and Transformers such as Pythia using Sparse Autoencoders. Across a 10‑million token corpus, 99.98% of Mamba features align closely with Pythia’s, supporting the Universality Hypothesis that core semantic representations are similar across architectures. A tiny 0.02% of features diverge, with Mamba’s recurrent bottleneck causing it to compress syntactic anomalies into polysemantic neurons, whereas Pythia’s attention can isolate distinct formatting edge‑cases.
The paper introduces EmoDiPyraTrans, a development-regularized differential graph Transformer designed to decode emotions from EEG signals across unseen individuals and populations while maintaining neural interpretability. Evaluations on five datasets (SEED, FACED, MAHNOB-HCI, DEAP, DREAMER) show high cross‑subject accuracies, with the model outperforming others in accuracy and positive‑class F1. Additional experiments demonstrate its ability to distinguish healthy from depressed participants and reveal spatial‑spectral neural signatures across frontal, temporal, central, and parietal regions with an alpha‑centered frequency preference.
The study audits demographic leakage in German-language resumes generated by large language models. Using ChatGPT, Gemini, and Qwen 3 variants, the authors generate resumes from anonymized profiles, varying only gender- and ethnicity-associated names while keeping qualifications constant. Even after anonymization and gender-neutralization, classifiers can reliably distinguish male- from female-generated resumes, driven by subtle differences in gender-neutral terminology rather than overtly gendered wording; ethnicity-related leakage remains weak.
The paper demonstrates that quantum transformer blocks can be intrinsically interpretable by tracking quantum mutual information, entanglement entropy, and state fidelity across layers. Experiments on four synthetic tasks show that learned mutual information aligns with task structure, entanglement is essential for accuracy, and mutual information predicts prediction correctness. These findings are validated on IBM Quantum hardware, illustrating that quantum computation’s physics can provide observable interpretability signals.
The paper introduces a pipeline and conversational system that processes 22,788 YouTube transcript and comment chunks from 309 North American cities to analyze public discourse on urbanism. It combines geographic entity resolution, topic modeling, sentiment analysis, and Retrieval-Augmented Generation (RAG), and reports empirical findings on model performance, such as a Twitter-tuned RoBERTa classifier outperforming VADER and dense retrieval surpassing TF‑IDF. The study also evaluates groundedness metrics, noting limitations of BERTScore and ROUGE‑1 for short user-generated text.
Euston is an 8‑B parameter mathematical claim‑verification model that resists producing false derivations when presented with corrupted theorems. It was trained on 3,026 matched true/corrupted statement pairs generated by GraphSynth, a probabilistic factor‑graph generator, and fine‑tuned from DeepSeek‑R1‑8B using GRPO. On a balanced held‑out split, Euston’s balanced accuracy rose from 29.50 % to 63.75 %, and its discrimination gap improved from –0.5 % to +27.5 %, while maintaining comparable general mathematical ability and reducing response length and truncation rates.
The paper presents a reproducible single‑subject baseline for reconstructing visual stimuli from EEG using a temporal‑spatial convolutional encoder that maps averaged EEG signals to 512‑dimensional ViT-B/32 image features. On the THINGS‑EEG2 dataset, the model achieves 12.83%, 39.17%, and 58.00% image recall at ranks 1, 5, and 10, respectively, outperforming analytical chance levels. The study also shows that performance drops sharply when applying a model trained on one subject to others, and that direct conditional generators without external visual weights produce noise‑dominated outputs, indicating that only coarse semantic decoding is feasible under the tested protocol.