arXiv cs.AI
Quantum Logic as the Logic of Contexts
We argue for the opposite order of explanation in a finite and fully computable setting. The free orthomodular lattice…
Hub / Ai Research
ai-research · WeSearch
Ai Research news.
Page 2 of Ai Research headlines on WeSearch — deduped and updated continuously from 10+ editorial sources.
arXiv cs.AI
Evolutionary Intelligence for Scientific Discovery: From Evolutionary Computation to Cumulative Discovery Systems
Evolutionary computation (EC) provides a computational basis for feedback-driven discovery because population-based…
arXiv cs.AI
Video Generation Models are General-Purpose Vision Learners
What, then, is the equivalent catalyst needed to achieve a general-purpose model in computer vision? In this paper, we…
arXiv cs.AI
Phone Segmentation and Recognition through Phonological Activation Mapping
Mortensen View a PDF of the paper titled Phone Segmentation and Recognition through Phonological Activation Mapping,…
arXiv cs.AI
Correlation-Aware Contextual Bandits with Surrogate Rewards for LLM Routing
Unlike classical contextual bandits that rely solely on bandit feedback and assume conditional independence across…
arXiv cs.AI
Model Agnostic Graph Prompt Learning for Crystal Property Prediction
These models often encode domain-specific knowledge into their graph encoding modules, which increases their parameter…
arXiv cs.AI
AlphaZero in Sparsely Rewarded Games: Limits and Auxiliary Supervision
We study this gap in two oracle-evaluable domains with contrasting structure: Connect Four, a solved partisan game…
arXiv cs.AI
SCATE: Learning to Supervise Coding Agents for Cost-Effective Test Generation
Currently, mitigating this premature termination requires continuous human-in-the-loop supervision. This heavy…
arXiv cs.AI
The Patchwork Problem in LLM-Generated Code
Computer Science > Software Engineering arXiv:2607.08981 (cs) [Submitted on 9 Jul 2026] Title:The Patchwork Problem in…
arXiv cs.AI
CLAP: Direct VLM-to-VLA Adaptation via Language-Action Grounding
Computer Science > Robotics arXiv:2607.08974 (cs) [Submitted on 9 Jul 2026] Title:CLAP: Direct VLM-to-VLA Adaptation…
arXiv cs.AI
MultiView-Bench: A Diagnostic Benchmark for World-Centric Multi-View Integration in VLMs
We introduce MultiView-Bench, a diagnostic benchmark expressly designed to evaluate multi-view integration for…
arXiv cs.AI
NL-PAC: Specification Ambiguity and Certified Minimax Risk Floors in LLM-Mediated Supervision
When a specification admits multiple readings but the supervision channel does not reveal which is operative,…
arXiv cs.AI
Eluna: An Agentic LLM System for Automating Warehouse Operations with Reasoning and Task Execution
We present Eluna, a production-deployed agentic system for reliable SOP execution. Eluna is a graph-guided,…
arXiv cs.AI
A Novel Parallel QCNN Architecture with Efficient Classical Simulability
Using a novel architecture inspired by previous QCNN and classical convolutional neural network (CNN) implementations,…
arXiv cs.AI
Prompt-Driven Exploration
Standard methods inject stochasticity in the action space, but such jitter only yields rollouts close to the original.…
arXiv cs.AI
TheBioCollection: Unified Pre-Training Scale LLM Corpus for Biology
However, existing biological resources, such as molecular databases, protein repositories, genomic annotations,…
arXiv cs.AI
Multi-Conditioned Diffusion Synthesis of Sand Boils for Low-Resource Earthen-Levee Inspection
We present a diffusion-based synthesis pipeline for low-resource sand-boil imagery. Using Stable Diffusion XL…
arXiv cs.AI
EHR-MPC: Inference-Time Control for Sepsis Treatment with Generative Patient Digital Twins
Existing reinforcement learning (RL) approaches learn fixed strategies for sepsis treatment, limiting adaptability to…
arXiv cs.AI
LLM-Driven Evolutionary Generation of Multi-Objective Bayesian Optimization Algorithms
We extend the LLaMEA framework to MOBO, using large language models as mutation and crossover operators within…
arXiv cs.AI
Accelerating GPU Inference of Large Language Models with Moderately Unstructured Sparse Weight Matrices
Pruning techniques that introduce sparsity into weight matrices can accelerate inference. However, maintaining model…
arXiv cs.AI
DaDaDa: A Dataset for Data Pricing in Data Marketplaces
Recognizing the value of data, data transactions are increasingly common, giving rise to many data marketplaces, e.g.,…
arXiv cs.AI
HERO: A Heterogeneity-Aware Benchmark Library for Federated Continual Learning
Computer Science > Machine Learning arXiv:2607.08784 (cs) [Submitted on 13 Jun 2026] Title:HERO: A Heterogeneity-Aware…
arXiv cs.AI
LieBN: Batch Normalization over Lie Groups
Recent advances have extended Deep Neural Networks (DNNs) to operate on manifolds, accompanied by normalization…
arXiv cs.AI
Director: Accelerating Distributed MoE Serving via Online Proactive Expert Placement
Its efficiency depends on the communication and computation latencies of the GPUs, which are linked to the placement…
arXiv cs.AI
Reward Transport: Property Control in Flow Matching via Noise-Space Alignment
We show that this coupling can instead serve as an alignment interface: by matching noise and data according to a…
arXiv cs.AI
Sticky Routing: Training MoE Models for Memory-Efficient Inference
Existing remedies are either system-level (caching heuristics) or post-hoc (router fine-tuning), leaving the root…
arXiv cs.AI
Signed Symmetric Quantization for Few-Bit Integers
Computer Science > Machine Learning arXiv:2607.08779 (cs) [Submitted on 12 Jun 2026] Title:Signed Symmetric…
arXiv cs.AI
iLENS: Interpretable LLM-Guided Mixture-of-Experts for Neuroimaging Survival Analysis
Predicting AD conversion during the prodromal stage remains critical for disease understanding and patient care. As…
arXiv cs.AI
A Unified Approach to Interpreting Knowledge Distillation for Large Language Models via Interactions
In this paper, we propose a unified approach to explore the common mechanism of various KD methods using interactions.…
arXiv cs.AI
REFORGE: A Method for Benchmarking LLMs' Reverse Engineering Capabilities in Decompiled Binary Function Naming
Computer Science > Software Engineering arXiv:2607.07738 (cs) [Submitted on 7 Jul 2026] Title:REFORGE: A Method for…
arXiv cs.AI
Minimal Decision Dynamics and Contextual Probability: A Quantum Tug-of-War Model
This paper develops a quantum-like extension of the Tug-of-War (QTOW) decision-making model to clarify when such…
arXiv cs.AI
ConceptSMILE: Auditing the Trustworthiness of Concept-Based Explainable AI
We introduce ConceptSMILE, a model-agnostic perturbation-based auditing framework for evaluating the reliability of…
arXiv cs.AI
Agora: Enhancing LLM Agent Reasoning Via Auction-Based Task Allocation
However, existing frameworks typically call APIs based on coarse-grained matching between tasks and the functions of…
arXiv cs.AI
TrustX Agent Risk Classification Framework (ARC): Risk-Tiering Internally Created Agentic AI Systems
Computer Science > Artificial Intelligence arXiv:2607.09586 (cs) [Submitted on 10 Jul 2026] Title:TrustX Agent Risk…
arXiv cs.AI
Knowledge Graphs and Explainable AI as Complementary Resources for Urban Mining
The relevant unit of value is not prediction accuracy alone, but the defensibility of the supported decisions: their…
arXiv cs.AI
Beyond Fixed Representations: The Vocabulary and Verifier Gaps in Open-Ended AI
These are powerful capabilities, but they share a structural limitation: the representational frame within which the…
arXiv cs.AI
SAGEAgent: A Self-Evolving Agent for Cost-Aware Modality Acquisition in Multimodal Survival Prediction
In multimodal clinical oncology, diagnostic modalities follow a clinically mandated order of escalating burden -- from…
arXiv cs.AI
Shared Selective Persistent Memory for Agentic LLM Systems
Naively persisting entire conversation histories is token-inefficient and counterproductive: irrelevant context…
arXiv cs.AI
Multimodal Reward Hacking in Reinforcement Learning
This risk is amplified when visual evidence is evaluated by text-only or weakly grounded rewards. We study reward…
arXiv cs.AI
Ceci n'est pas une pipe: AI systems as semantic abstractions
We propose a semantic framework to describe AI systems, to be able to examine the correctness of such representations.…
arXiv.org
NormAct: A Benchmark for Hidden Social Norm Compliance in Embodied Planning
While explicit goals may render certain actions optimal, implicit social norms often impose hidden constraints.…
arXiv.org
ATOD: Annealed Turn-aware On-policy Distillation for Multi-turn Autonomous Agents
On-policy distillation (OPD) provides dense teacher guidance and typically improves rapidly in the early stage, but…
arXiv.org
Internalizing the Future: A Unified Agentic Training Paradigm for World Model Planning
The paper introduces a unified training paradigm that equips large language model agents with internal world modeling…
arXiv.org
Towards Reliable and Robust LLM Planning: Symbolic Feedback-Driven Iterative Self-Refinement Framework
Planning, a core component of intelligent behavior, remains challenging for LLMs, which often produce infeasible or…
arXiv.org
ToE: A Hierarchical and Explainable Claim Verification Framework with Dynamic Multi-source Evidence Retrieval and Aggregation
The paper presents Tree of Evidence (ToE), a hierarchical framework for automated claim verification that builds…
arXiv.org
MER-R1: Multimodal Emotion Reasoning via Slow-Fast Thinking Synergy
Specifically, for reasoning-based MLLMs, fast thinking by triggering direct answers often outperforms slow thinking…
arXiv.org
DysLexLens: A Low-Resource LLM Framework for Analysing Dyslexic Learners Insights from Online Forums
However, their lived experiences with these tools remain largely underexamined. This paper proposes DysLexLens, a…
arXiv.org
AI-Model Network: Concept, Current State and Future
Computers create the Internet, and the Internet empowers the value of computers. The rapid development of the…
arXiv.org
When Does Personality Composition Matter for Multi-Agent LLM Teams?
Computer Science > Artificial Intelligence arXiv:2606.27443 (cs) [Submitted on 25 Jun 2026] Title:When Does…
arXiv.org
Odyssey: Constructing Verifiable Local Truth-Preserving Foundation Models
A foundry is an organized sheaf of knowledge that carries within it an argumentation component. Concrete foundries are…