Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
Ensuring reliability in uncertain environments remains difficult for long-horizon robotic manipulation. End-to-end VLA models are data-heavy and opaque, making diagnosis and verification difficult. Hi...
Large-scale rollouts have become a core component of modern LLM systems, spanning reinforcement learning (RL) post-training, on-policy distillation (OPD), and sampling-heavy evaluation pipelines. Unli...
arXiv:2501.10573v2 Announce Type: replace-cross Abstract: We study the geometry of token representations at the prompt level in large language models through the lens of intrinsic dimension. Viewing...
RouteScan is a non‑intrusive auditing framework that detects harmful behavior in Mixture‑of‑Experts (MoE) large language models by analyzing expert‑routing telemetry captured from GPU execution. It uses the number of active GPU threads during the prefilling phase as a micro‑architectural fingerprint to isolate cross‑domain risk indicators and precisely identify malicious prompts. Evaluations on four open‑source MoE LLMs show strong generalization with AUROC > 0.91 on unseen harmful domains, while privacy tests indicate that full prompts cannot be reliably recovered from aggregated telemetry.
arXiv:2608.21023v1 Announce Type: new Abstract: Word alignment has traditionally been studied between sentences, but many cross-lingual tasks increasingly require correspondences across full document...
arXiv:2608.20430v1 Announce Type: new Abstract: World Action Models (WAMs) improve planning by incorporating future world evolution into action generation, yet existing methods allocate a fixed imagi...
arXiv:2602.17850v2 Announce Type: replace-cross Abstract: Conversational agents increasingly mediate everyday digital interactions, yet the effects of their communication style on user experience and...
arXiv:2608.21300v1 Announce Type: new Abstract: Foundation models for medical image segmentation, like prompt-based MedSAM, generalize well across domains and modalities, often in zero or few-shot se...
arXiv:2608.20999v1 Announce Type: new Abstract: Multimodal LLMs apply the language model interface to visual inputs, where ordinal regression tasks such as age estimation, image quality assessment, a...
arXiv:2608.10725v2 Announce Type: replace Abstract: Large language models (LLMs) often rely on shortcuts rather than systematic reasoning, raising safety concerns in medical applications. Allowing mo...
MatMMExtract is an open‑source pipeline that disassembles compound scientific figures into individual sub‑panels and generates structured, grounded image‑text pairs using a large language model guided by a materials science taxonomy. Applied to 14,810 open‑access articles, it produced 391,606 panel‑level pairs with sub‑captions, a two‑level visualisation category (19 classes, 100+ subtypes), and scientific summaries. The project also introduces MaterialScope, a 2,811‑figure detection dataset, and demonstrates that Gemini 3.1 Flash Lite yields high‑quality annotations with low hallucination, while a dual‑encoder baseline outperforms zero‑shot CLIP on the resulting MatSciFig dataset.
arXiv:2608.11645v2 Announce Type: replace Abstract: Volumetric video streaming turns privacy into a 3D, multi-view problem. Unlike ordinary video, where sensitive content can often be redacted frame...
The paper introduces a new framework for identifying average dose-response functions in the presence of unmeasured confounding by using instrumental variables. It defines a uniform regular weighting function and partitions the treatment space into open sets where local identification is possible. For estimation, the authors propose an augmented inverse probability weighted score within a debiased machine learning setting, along with practical guidance for constructing weighting functions, falsification tests for the additive IV condition, and asymptotic theory for kernel regression or empirical risk minimization estimators.
arXiv:2608.20610v1 Announce Type: cross Abstract: We study causal inference in interrupted time series designs where a treatment affects every unit simultaneously, so that the contemporaneous control...
The paper presents an end‑to‑end framework for extracting and clustering trilingual Sri Lankan parliamentary debates in Sinhala, Tamil, and English. Using LLM‑based text extraction, multilingual embeddings, and density‑based clustering, the authors recover 30 macro‑topics with a cluster purity of 0.673. The temporal patterns of these topics align with major national events such as the 2019 Easter attacks and the 2022 economic crisis, demonstrating the method’s effectiveness where traditional LDA fails.
arXiv:2605.19624v3 Announce Type: replace-cross Abstract: For camera-based satellite visual sensing, Sim2Real data construction requires images that approach real-domain sensor appearance while retai...
arXiv:2608.20710v1 Announce Type: new Abstract: Real-world semi-supervised learning (SSL) often encounters significant challenges with long-tailed label distributions and noisy pseudo-labels, which h...
COEC (Calibrated Orthogonal-Equivalence Compensation) is a training‑free framework that improves structured pruning of large language models by applying alternating left and right orthogonal rotations to the retained weight matrix. The method optimizes the right rotation on a reduced Stiefel manifold, rescales singular values via generalized cross‑validation, tempers the calibration Gram matrix, and adds an alignment penalty to preserve geometric relations between attention projections. Experiments on Llama‑3, Llama‑3.1, and Qwen2.5 show that COEC consistently improves perplexity and zero‑shot accuracy across multiple sparsity levels, outperforming existing compensation techniques.
arXiv:2608.21133v1 Announce Type: new Abstract: Medical image-text data can expose protected health information (PHI) through both visible image content as well as accompanying text, creating a barri...
Explainable Deepfake Detection with Feature-robust Augmentation and Evidence-grounded Explanation Optimization proposes a new framework that improves deepfake detection and interpretability. The approach introduces Feature-robust Augmentation—diversified degradation-aware strategies combined with supervised contrastive learning and a mean-teacher architecture—to maintain accuracy on low-quality images. For explanations, it employs evidence-grounded preference optimization, guiding the model to focus on genuine manipulation traces by learning from chosen-rejected explanation pairs that omit evidence or inject irrelevant details. The method achieved first place in the ACM Multimedia 2026 Explainable Deepfake Detection Challenge and is publicly available on GitHub.