arXiv:2605.27328v1 Announce Type: cross Abstract: Recent advances in agentic systems increasingly treat code as an executable operational substrate rather than as a disposable output artifact. Prior work such as emphCode as Agent Harness frames validated agent-generated artifacts as runtime entities that can be created, executed, revised, persisted, and reused within long-running cognitive loops. However, the […]
Curriculum Learning for Safety Alignment
arXiv:2605.26315v1 Announce Type: cross Abstract: Direct Preference Optimisation (DPO) is widely used for safety alignment in large language models. However, prior work shows it is brittle and exhibits poor out-of-distribution (OOD) generalisation. In this paper, we investigate whether Curriculum Learning can improve the robustness of DPO-based safety alignment. We propose Staged-Competence, a curriculum-based framework that […]
A Universal Cliff and a Design Fingerprint: Cross-Section Defect Detection Under LLM Orchestration
arXiv:2605.26174v1 Announce Type: cross Abstract: Production language-model systems answer a request by partitioning it across an invisible orchestration of worker agents that recompose one integrated report. We ask what this does to a class of defect no single worker can see: a contradiction in the relation between two distant sections of a document. Holding the […]
HRVConformer: Neonatal Hypoxic-Ischemic Encephalopathy Classification from the Heart Rate signals
arXiv:2605.26190v1 Announce Type: cross Abstract: This paper presents the HRVConformer, a novel deep learning architecture for the classification of hypoxic-ischemic encephalopathy (HIE) using the instantaneous heart rate (HR) signal. Unlike conventional approaches that rely on handcrafted features, HRVConformer directly processes raw HR signals in an end-to-end manner, capturing both local and long-range dependencies through a […]
Edge AI Deployment Beyond Models: A BSP-Aware Systems Framework for Industrial Embedded Platforms
arXiv:2605.26119v1 Announce Type: cross Abstract: Industrial Edge AI programs often begin with the model and only later confront the platform. That sequencing is attractive because it allows early demonstrations, but it breaks down when the deployment target is an embedded system with long product lifecycles, vendor-specific kernels, heterogeneous accelerators, safety constraints, and nontrivial I/O paths. […]
When Does Adaptive Guidance Help? Belief-Aware Privileged Distillation for Autonomous Driving Under Partial Observability
arXiv:2605.26155v1 Announce Type: cross Abstract: Guided Soft Actor-Critic (GSAC) distills knowledge from a privileged full-state teacher to a partial-observation student for autonomous driving, but uses a fixed distillation coefficient lambda regardless of the agent’s uncertainty. We present Belief-Aware GSAC (BA-GSAC), which modulates lambda via ensemble disagreement, and use it as a testbed for a systematic […]
ICCU: In-Context Continual Unlearning via Pattern-Induced Refusal Rules
arXiv:2605.27138v1 Announce Type: new Abstract: Machine unlearning aims to remove the influence of specific data from trained language models. In real-world deployments, unlearning requests often arrive sequentially, which challenges existing fine-tuning-based methods: fine-tuning each request is costly, accumulates utility loss, and may cause cross-request interference. To address these issues, we propose ICCU (In-Context Continual Unlearning), […]
Query Symbolically or Retrieve Semantically? A Dataset and Method for Semi-Structured Question Answering
arXiv:2605.27164v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) systems for question answering typically retrieve evidence by semantic similarity between the query and document chunks. While effective for unstructured text, this approach is less reliable on semi-structured corpora where answering may require exact filtering, aggregation, or exhaustive retrieval over structured attributes across multiple documents. Symbolic approaches […]
Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases
arXiv:2605.27355v1 Announce Type: new Abstract: Reinforcement Learning from Human Feedback (RLHF) is the standard method to align Large Language Models (LLMs) with human preferences. In this work, we introduce alignment tampering, a potential vulnerability where the LLM undergoing alignment influences the preference dataset, causing RLHF to amplify undesired behaviors. This arises from core limitations of […]
AssetGen: Deployable 3D Asset Generation at Interactive Speed
arXiv:2605.26137v1 Announce Type: cross Abstract: While 3D generation is progressing rapidly, recent work has often focused on obtaining high-resolution assets, leaving user experience and deployability as afterthoughts. We present AssetGen, a 3D generator that focuses instead on these two aspects. Given one reference image, in 30 seconds it produces a high-quality mesh with baked normals, […]
On the Push-Based Asynchronous Federated Learning: A Bias-Correction Aggregation Approach
arXiv:2605.26162v1 Announce Type: cross Abstract: Asynchronous decentralized federated learning (ADFL) eliminates central coordination and global synchronization, making it attractive for large-scale and heterogeneous systems. However, frequent peer-to-peer communication, asynchronous updates on directed topologies, and non-IID data jointly lead to excessive communication overhead, biased aggregation and severe model drift. We propose PushCen-ADFL, a communication-efficient ADFL framework […]
Position: AI Safety Requires Effective Controllability
arXiv:2605.27117v1 Announce Type: new Abstract: AI safety is still largely framed as alignment: training models to follow human preferences, safety policies, and normative constraints. That framing has improved the behavior of modern language models, but aligned behavior does not by itself guarantee that a deployed agent can be stopped, overridden, or constrained once it operates […]