Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv:2608.28929v1 Announce Type: cross Abstract: Large-scale diffusion models have fueled numerous profitable downstream applications for AI-related businesses, including visual editing and content...
arXiv:2505.22850v3 Announce Type: replace Abstract: Referring Expression Counting (REC) requires distinguishing visually similar objects described by fine-grained text cues. Existing methods tackle t...
arXiv:2601.03468v2 Announce Type: replace Abstract: Reinforcement learning (RL) has become a standard approach for post-training large language models and, more recently, for improving image generati...
arXiv:2604.22280v4 Announce Type: replace Abstract: Multimodal Large Language Models (MLLMs) have emerged as a promising foundation for universal multimodal embeddings. Recent studies have shown that...
arXiv:2605.09687v2 Announce Type: replace Abstract: Remote sensing single-image super-resolution aims to generate high-resolution imagery from low-resolution observations while preserving fine struct...
arXiv:2605.22311v2 Announce Type: replace Abstract: Identity-conditioned diffusion models enable high-quality and identity-consistent face generation, but they also raise severe privacy concerns, as...
arXiv:2606.22394v3 Announce Type: replace Abstract: Consistency distillation has significantly accelerated diffusion-model inference, but its sampling dynamics remain underexplored. We reveal an asym...
arXiv:2607.16862v2 Announce Type: replace Abstract: LiDAR place recognition supports loop closure, relocalization, and multi-agent map management. As robotic platforms increasingly combine LiDARs wit...
arXiv:2311.08139v2 Announce Type: replace-cross Abstract: Feedforward neural networks (FNNs) are typically viewed as pure prediction algorithms, and their strong predictive performance has led to the...
SafeAtlas-VL introduces a large multimodal safety dataset with 1.5 million instances, rating image, request, and response risks on a five‑level ordinal scale across 15 harm categories and 55 subcategories. The accompanying SafeAtlas‑Bench provides 5,000 held‑out cases for evaluating ordinal predictions and continuous risk scores. Models trained on this data, including an 8B Guard model, achieve state‑of‑the‑art performance, outperforming prior benchmarks by about 4% in F1 score.
CoLa-ICD is a knowledge‑enhanced framework designed to improve automatic medical coding of ICD codes in long, imbalanced clinical documents. It enriches ICD labels with external terms, models dependencies among related codes, and strengthens the alignment between label semantics and clinical evidence, particularly for rare codes. Experiments demonstrate that CoLa-ICD achieves state‑of‑the‑art performance in AUC, F1, and P@k, with larger gains in larger and sparser label spaces.
The paper introduces TPR-Attention, an attention mechanism that operates over tensor‑product representations to embed structured inductive bias into deep learning models. Experiments on compositional tasks demonstrate that TPR‑Attention outperforms existing architectural components in achieving combinatorial generalization. The results suggest that incorporating explicit compositional structure into neural attention can improve systematic generalization.
The paper introduces CHAP, a personalized generative retrieval framework that aligns query semantics with item representations through a hierarchical semantic alignment module and models user behavior using both discrete Semantic IDs and continuous representations. It also proposes a Residual Cascading Generation mechanism to reduce inference latency by limiting the Transformer decoder to a single pass. Experiments on multiple datasets and online A/B tests show that CHAP outperforms existing methods, demonstrating its practical value.
The paper proposes reverse n‑wise output‑oriented testing for AI/ML and quantum computing systems, a method that builds covering arrays over output equivalence classes, confidence buckets, decision boundaries, fairness partitions, embedding clusters, ranking stability bands, quantum measurement distributions, and error syndrome patterns. It then uses gradient‑free metaheuristic optimization to solve the inverse mapping problem, generating input configurations or quantum circuit parameters that trigger specific behavioral signatures in opaque models. The framework claims to provide explicit coverage guarantees, higher fault detection rates for calibration, boundary, and error syndromes, improved test suite efficiency, and automated partition discovery for MLOps and quantum validation pipelines.
The paper introduces TrustFlip, an attack that exploits consistency‑based defenses in vehicular collaborative perception by deploying physical adversarial objects to create inconsistent observations among benign vehicles. This misattribution lowers the trust score of a targeted vehicle, leading to its exclusion from the collaboration and a degradation of perception performance. The authors evaluate the attack across multiple architectures, showing it can remove a benign vehicle in up to 87.7% of scenarios and reduce Average Precision by up to 13%, and propose a mitigation called TrustReflect that reduces the attack success rate by 35–100%.
The study investigates how large language models (LLMs) assess psychological distress in online posts from six identity‑based communities. Through a perspectivist annotation task, 321 participants provided 9,587 judgments on 1,198 Reddit posts, revealing modest in‑group agreement (OR = 1.18) that varies across communities. When evaluated against these community‑specific labels, open‑weight LLMs consistently over‑estimate distress—achieving only 31–44% accuracy on posts perceived as none‑to‑mild—while newer models like GPT‑5 and Gemini 2.5 Pro show similar inflation, whereas Claude Opus 4 is more conservative. "whyItMatters":"The findings highlight that miscalibrated distress detection by LLMs can disproportionately impact the very communities they aim to serve, underscoring the need for equitable AI deployment in mental‑health contexts."
The paper introduces ACTD, an Anchor-Based Cross-Tokenizer Distillation method that aligns vocabularies and sequences to transfer reasoning capabilities from large language models to smaller students. It uses a novel anchor loss with residual regularization to reduce alignment noise and extends the approach to multiple teachers. Experiments on five reasoning benchmarks with three teachers show state‑of‑the‑art results, with the multi‑teacher variant outperforming existing baselines.
arXiv:2608.29208v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models, built upon Vision-Language Models (VLMs), have significantly enhanced robotic capabilities by leveraging interne...
arXiv:2608.30437v1 Announce Type: new Abstract: Graph-augmented large language models often assume that graph evidence produced by external computation and placed in the input can be used by the nati...
arXiv:2608.29270v1 Announce Type: cross Abstract: Autoformalization translates informal mathematical theorems into code for proof assistants such as Lean. A central challenge is that current evaluati...