SchemaWeaver: Accelerating On-Device Mobile Agentic AI via Schema-Guided Parallel LLM Decoding

Boyuan Yang, Wei Gao

#arXiv #cs.AI

Agency as a New Organizing Abstraction for Computer Science

Ian Foster

#arXiv #cs.AI #cs.DC #cs.HC #cs.MA

When Goals Leak Across Relationships: How Cached Goal Representations Shape Cross-Client Decisions

Saloni Gupta, S\'ebastien Krier, Glenn Matlin

#arXiv #cs.AI

RL Post-Training Diffusion Language Models Along Their Own Generation Paths

Vaibhav Singh, Eugene Belilovsky

#arXiv #cs.AI

ChemExplain: A Multimodal Benchmark for Explanatory Chemical Reasoning

Aotian Chen et al.

#arXiv #cs.AI #cs.CV

📌 Understanding Personalization Gains in Whisper

Sergei Katkov, Antonio Liotta

#arXiv #cs.AI #cs.CL #cs.LG

Neural Bayesian Networks for Modular and Scalable Decision-Making Under Uncertainty

Giovanni Briglia et al.

#arXiv #cs.AI #cs.LG

Observe, Share, or Predict? Disentangling Partner Information in Embodied Multi-Agent Collaboration

Adri\'an Szlatincs\'an et al.

#arXiv #cs.AI #cs.RO

📌 When 1+1 Backfires: Rethinking Relevance-Diversity Balancing for Sample-Aware Visual Token Reduction

Yihang Huang et al.

#arXiv #cs.AI

XAI-Guided Fourier Domain Generalisation for Pure-Simulation Sim-to-Real Deep Reinforcement Learning

Wei Yao et al.

#arXiv #cs.AI #cs.CV

FON: A Draft-Revise Dual-Organ Language Model without Sequence-Level Quadratic Attention

Helong Hu (Zhanjiang Campus et al.

#arXiv #cs.AI

REVEAL: Reverse Engineering Agentic Latents from Frontier-Model Traces

Yuetian Du et al.

#arXiv #cs.AI #cs.CL

Learning to GUIDE the Group: Self-Learning LLM Facilitation Across Multi-Party Tasks

Ziyi Liu et al.

#arXiv #cs.AI

Correct Choices, Failed Sequences: Choice2Seq Benchmarks Correctness Retention Across Output Interfaces

Tianyou Lai et al.

#arXiv #cs.AI

Interactive 4D Volumetric Liquid Forecasting under Moving-Solid Interaction

Jongwon Ryu, Sunjae Yoon, Junyeong Kim

#arXiv #cs.AI #cs.CE #cs.LG

Agent Sociodrama: Social-Dynamics-Guided Persona Recasting for LLM Agent Teams

Yirui Qi, Xiaoming Zhang, Jaehong Yoon

#arXiv #cs.AI #cs.CL #cs.MA

Data Portability Revisited: Toward the Human-Centric, AI-Driven Data Ecosystems of Tomorrow

Mark Fenwick et al.

#arXiv #cs.AI

Running Fewer Loops Removes Refusal Behaviour in Recurrent-Depth Language Models

Seyedarmin Azizi et al.

#arXiv #cs.AI

Harmony Harness: Multi-Objective Self-Evolution of Agent Harnesses

Weichen Yu et al.

#arXiv #cs.AI

Trust Hinders Bidirectional Human-AI Alignment

Michaela Benk, Kacper Sokol

#arXiv #cs.AI #cs.HC