arXiv:2608.23001v1 Announce Type: new Abstract: Automated manuscript pipelines often regenerate an entire section to repair a local defect, allowing…
Category: cs.AI updates on arXiv.org
PsychJail: Exploring Psychological Jailbreaks via Multi-Turn Persuasion of LLM Policies
arXiv:2608.23028v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed in education, healthcare, policy advising, and…
MobilePA-Bench: Benchmarking Mobile Planner Agents on Complex Real-World Tasks
arXiv:2608.23035v1 Announce Type: new Abstract: As on-device LLM agents evolve into personal copilots, the mobile operating system has become a key…
Budget-Constrained Embodied Perception: Four Resource Walls and a Pre-Registered Evaluation of Access-Structured Perception on Open Models at less than 31B
arXiv:2608.22975v1 Announce Type: new Abstract: Embodied multimodal agents must answer from growing observation streams under a fixed per-decision token…
Buried in Textual Debt: Context Pruning with Visual Evidence Preservation for MLLM Agents
arXiv:2608.22963v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) are increasingly deployed as multi-step agents, where explicit…
SA-RSQ: A Versatile Sparse Representation Framework for Multi-modal Recommender Systems
arXiv:2608.22979v1 Announce Type: new Abstract: Deploying high-dimensional multimodal features in industrial recommender systems incurs substantial…
Toward Effective and Reliable LLM Agents via Dynamic Ontology
arXiv:2608.22974v1 Announce Type: new Abstract: Large language model (LLM) agents rely heavily on knowledge encoded in model parameters or presented as…
ParallelWorld: Test-Time Scaling for Embodied Reasoning
arXiv:2608.22971v1 Announce Type: new Abstract: Embodied Reasoning constitutes a fundamental capability of embodied intelligence, serving as the basis for…
CDEG: Learning Decision-Critical Evidence for Long-Horizon Diagnostic Agents
arXiv:2608.22899v1 Announce Type: new Abstract: Unlike static medical question answering, long-horizon diagnosis captures the sequential nature of…
Concepts for Securing Agentic AI Coding and the Terok Environment
arXiv:2608.22930v1 Announce Type: new Abstract: Agentic AI is a fascinating new tool for software development. It is a huge step forward compared to…