arXiv AI
Jul 14
Dynamic Agent Skills: A Lifecycle Survey and Taxonomy of Evolving Skill Libraries
arXiv:2607. 10113v1 Announce Type: new Abstract: Large language model agents increasingly store reusable procedures outside the model.
AI safety and alignment
Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv AI
Jul 14
Pipette: An Embodied Simulation Platform, Benchmark, and Data-Efficient Augmentation Framework for Wet-Lab Robotics
arXiv:2606. 12936v2 Announce Type: replace-cross Abstract: Wet-lab robots can improve the reproducibility, throughput, and safety of biomedical experiments, but scaling their learning requires customizable simulators for safe and reproducible task generation, open editable laboratory assets, and efficient pipelines that turn limited demonstrations into usable training data.
arXiv Machine Learning
Jul 14
Generative Testing of Automated Speech Recognition Systems
arXiv:2607. 09833v1 Announce Type: cross Abstract: Automatic speech recognition (ASR) systems have achieved high accuracy with transformer-based models, enabling deployment in critical applications.
arXiv AI
Jul 14
Workload-Driven Optimization for On-Device Real-Time Subtitle Translation
arXiv:2607. 09957v1 Announce Type: cross Abstract: This report studies on-device English-to-Traditional-Chinese subtitle translation for Taiwan under short inputs, short outputs, batch-size-one inference, low latency, and privacy constraints.
arXiv Machine Learning
Jul 14
Selective Safety Steering via Value-Filtered Decoding
arXiv:2605. 14746v2 Announce Type: replace Abstract: While large language models (LLMs) are trained to align with human values, their generations may still violate safety constraints.
arXiv AI
Jul 14
Cost of Reasoning in non-English Languages: A Case Study on Japanese
arXiv:2607. 10114v1 Announce Type: cross Abstract: Reasoning Language Models (RLMs) achieve their strongest performance when they reason in English, the language for which reasoning-oriented training data is most abundant.
arXiv AI
Jul 14
Beyond Euclidean Clipping: Overcoming Exploration Collapse in LLM RL via Riemannian Isometric Policy Optimization
arXiv:2607. 10169v1 Announce Type: cross Abstract: Reinforcement learning (RL) has become a dominant paradigm for enhancing LLMs' reasoning capabilities.
arXiv Machine Learning
Jul 14
Quota Marketplace: Dynamic Pricing for Efficient Allocation of ML Training Resources
arXiv:2607. 09802v1 Announce Type: new Abstract: The escalating demand for Machine Learning (ML) training resources in recent years has resulted in a substantial gap between the high demand and the available supply.
arXiv AI
Jul 14
BackgroundMellow: A Multi-Modal Cohesive Framework for Narrative-Driven Rich Cinematic Soundscape Generation
arXiv:2607. 11364v1 Announce Type: cross Abstract: Generating immersive, synchronized and cinematic audio for long-form textual narratives remains a significant challenge in multi-modal AI.
arXiv Machine Learning
Jul 14
Normative Alignment of Recommender Systems via Internal Label Shift
arXiv:2607. 10915v1 Announce Type: cross Abstract: We introduce NAILS (Normative Alignment of Recommender Systems via Internal Label Shift), a simple and scalable method for aligning recommendation outputs with target distributions over item-level attributes, such as categories.
arXiv AI
Jul 14
Verifier-Guided Twelve-Tone Composition: A Generate-Verify-Repair Harness for Symbolic Music Generation
arXiv:2607. 11334v1 Announce Type: new Abstract: Large language models can produce superficially legal twelve-tone scores that collapse into degenerate textures.
arXiv AI
Jul 14
VehAnchor: Metadata-Free Metric Scale Recovery from Vehicle Cues in Aerial Imagery
arXiv:2603. 04277v2 Announce Type: replace-cross Abstract: Autonomous aerial robots operating in GPS-denied or communication-degraded environments frequently lose access to camera metadata and telemetry, leaving onboard perception systems unable to recover the absolute metric scale of the scene.
arXiv AI
Jul 14
MG$^2$-RAG: Multi-Granularity Graph for Multimodal Retrieval-Augmented Generation
arXiv:2604. 04969v2 Announce Type: replace-cross Abstract: Retrieval-Augmented Generation (RAG) mitigates hallucinations in Multimodal Large Language Models (MLLMs), yet existing systems struggle with complex cross-modal reasoning.
arXiv AI
Jul 14
Towards Predictive, Aligned, and Scalable Robot Learning
arXiv:2607. 11270v1 Announce Type: cross Abstract: Learning, at its core, extends beyond memorization to the ability to reason and solve novel problems by navigating a space of possibilities.
arXiv AI
Jul 14
Large Language Models in Misinformation Ecosystems: Misuse, Defense, and Vulnerability
arXiv:2607. 10402v1 Announce Type: cross Abstract: Large language models (LLMs) have transformed misinformation from a primarily content-centric problem into a broader ecosystem-level security challenge.
arXiv AI
Jul 14
What You Train Is What You Get: Gender Bias, Training Composition, and Post-Hoc Mitigation in Audio Deepfake Detection
arXiv:2607. 09891v1 Announce Type: cross Abstract: Audio deepfake detection models determine whether speech is genuine or artificially generated, but high overall accuracy can mask substantial performance disparities across demographic groups.
arXiv AI
Jul 14
When Are Sparse Feature Interventions Actually Localized? Matched Evaluation for SAE-Based Safety Control
arXiv:2607. 10226v1 Announce Type: new Abstract: We evaluate when sparse autoencoder (SAE) features act as localized control handles for safety-relevant behavior.
arXiv AI
Jul 14
On the Necessity of Output Distribution Reweighting for Effective Class Unlearning
arXiv:2506. 20893v5 Announce Type: replace-cross Abstract: In this paper, we reveal a significant shortcoming in class unlearning evaluations: overlooking the underlying class geometry can cause information leakage about the forgotten class.
arXiv AI
Jul 14
Diachronic Sample Integration: Robust Tail-Risk Estimation with Generative Models
arXiv:2607. 10810v1 Announce Type: cross Abstract: Deep generative models are increasingly used as simulators for downstream decision-making under data scarcity, but in risk-sensitive applications their usefulness depends on rare adverse scenarios rather than typical samples.
arXiv AI
Jul 14
Annotation-Free Furniture Codes: What They Encode, and How Far They Transfer
arXiv:2607. 10461v1 Announce Type: cross Abstract: Layout-based 3D scene synthesizers place each object using two human-annotated channels: a categorical class label and a canonical-pose convention.