arXiv:2609.13257v1 Announce Type: cross Abstract: Test-time scaling (TTS) can improve generation only when additional compute produces better candidates…
Category: cs.AI updates on arXiv.org
(How) Do MLLMs Report Bistable Images Like Humans?
arXiv:2609.13254v1 Announce Type: cross Abstract: Bistable images such as the duck-rabbit are classic stimuli in which one image supports multiple…
An Evolutionary Computation Framework for Multi-Agent Q-Learning with Mean-Field Environmental Feedback
arXiv:2609.13253v1 Announce Type: cross Abstract: Multi-agent reinforcement learning in networked populations is governed by the interaction between…
Evaluation of MLLM-Agnostic Plug-and-Play Keyframe Selection Methods for Long Video Understanding
arXiv:2609.13250v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) cannot process every frame of a long video because of…
LLMs or Naive Bayes? Old Gems or New Ways
arXiv:2609.13185v1 Announce Type: cross Abstract: Large language models (LLMs) prompt a recurring question in research computing: should classical methods…
Diagnosing Faults in Reinforcement Learning Simulators and World Models with Canonical Polynomial Invariants
arXiv:2609.13194v1 Announce Type: cross Abstract: A large literature builds physical structure into learned dynamics on the premise that models respecting…
ShieldVLA: Feasibility-Aware Safety Alignment for Vision-Language-Action Models
arXiv:2609.13231v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models demonstrate strong generalization in robotic manipulation and…
Chemical and geometric representation fidelity improves drug–target affinity prediction
arXiv:2609.13230v1 Announce Type: cross Abstract: Predicting drug–target binding affinity (DTA) requires models to distinguish subtle chemical and…
Planning as Dynamics Relaxation: Hippocampal Recurrent Network Realizes Optimal Goal-Directed Navigation
arXiv:2609.13219v1 Announce Type: cross Abstract: Neural correlates of spatial cognitive map are well documented, yet exactly how neural circuits perform…
Natural-Language to SysMLv2 Translation via Conformance-Driven Iterative Refinement
arXiv:2607.14162v1 Announce Type: cross Abstract: Model-Based Systems Engineering (MBSE) relies on formal system models as primary technical artifacts for…