314 npm packages were compromised in an attack attributed to Mini Shai-Hulud. The malicious packages contained data exfiltration code. The incident raises concerns about software supply chain security.
arXiv study analyzing 10,000+ Google Maps reviews of urgent care facilities (DMV, Florida) using GPT prompt engineering for aspect-based sentiment extraction. Findings: interpersonal factors and operational efficiency are strongest drivers of patient satisfaction; technical quality, finances, facilities show no significant independent effects. Population density alone shows modest correlation with ratings.
LERA introduces a two-stage ad auction framework for LLM chatbots: coarse filtering via embeddings, then fine ranking using LLM prompts for relevance scores. Combines organic relevance with bids using critical-value payment rule. Improves ad selection accuracy and insertion diversity on synthetic benchmark.
FedNL reformulates federated learning as a three-level nested optimization system. Embeds Titans-based linear attention to enable zero-shot test-time adaptation without additional training. Experiments on MMLU and long-context benchmarks show competitive performance and constant inference memory.
On-device AI system for ecological monitoring in remote areas. Architecture separates visual perception from reasoning using dynamic knowledge base, eliminating cloud dependency and continuous retraining. Collaboration with biologists and Indigenous communities for ethical AI co-development.
Automatic generation of fuzzy cognitive maps (FCMs) from text using LLM agents that chunk text into overlapping segments. Convex mixing of chunk FCMs produces a cyclic FCM knowledge graph. Operator-level Bayesian inference generates "de-chunked" FCMs. Demonstration on Thucydides Trap model: 7 out of 8 FCMs predicted armed conflict. Gemini 3.1 served as chunking agent.
Neural network weights constitute a first-class data modality. This position paper proposes treating checkpoints as generative data: synthesizing weights on demand matches or exceeds fine-tuning while reducing adaptation costs by orders of magnitude. High-performing models occupy structured regions of weight space (symmetry, modularity, shared subspaces).
Study using Transcranial Doppler (TCD) and MOCAIP algorithm to predict brain vascular age. 168 healthy and 277 diseased subjects (stroke, Alzheimer's, MCI) analyzed. Model predicts accelerated aging in patients, with 3.69-year overestimation in healthy subjects.
Systematic survey of deep learning architectures for 3D point cloud classification and segmentation. Addresses challenges (unordered nature, sensor noise, occlusions) and strategies (format conversion, local geometry extraction, permutation-invariant processing, self-attention). Evaluates models on standard benchmarks.
Investigation of how action information is incorporated into RNN state update functions for reinforcement learning. Authors empirically evaluate several recurrent cell architectures on illustrative domains and identify critical design choices often unreported in large-scale RL agents.
IsalProgram is a regular assembly-like language where every finite string is a valid program. Executed on a virtual machine with circular doubly linked list and three data pointers, it eliminates memory addresses and variable names. Proposed as a target for neural program synthesis.
arXiv study analyzing 10,000+ Google Maps reviews of urgent care facilities (DMV, Florida) using GPT and prompt engineering. Interpersonal factors and operational efficiency emerge as primary satisfaction drivers, while technical quality, finances, and facilities show no significant independent effects. Population density alone correlates with ratings among socioeconomic factors.
Theoretical paper on evaluating LLM-based agent-based models. Authors propose a four-level Mechanism Plausibility Scale to distinguish generative sufficiency (ability to reproduce phenomena) from mechanistic plausibility (how phenomena are produced), grounded in philosophy of science.
Decentralized task allocation framework for circular smart manufacturing using Edge-AI and learning-to-rank. Each machine evaluates incoming tasks with local information (processing capability, queue state, resource contention). Results: improved delay, better deadline adherence, enhanced energy efficiency in simulation.
Survey on integrating foundation models (LLMs, Gemini, Grok) with federated learning to enable artificial personalized intelligence (API). Proposes personalized federated intelligence (PFI) paradigm combining privacy, generalization, and edge personalization, trustworthy adaptation, and refinement via retrieval-augmented generation.
PersonaArena is a dynamic simulation framework for evaluating and improving persona-level role-playing in LLMs. It leverages a filtered corpus of user-generated social content, constructs a nuanced persona bank, and simulates multi-turn interactions in social environments. A multi-agent debating judge provides holistic and unbiased assessment.
Experimental study testing Chomsky's critique of LLMs: GPT-2 small and LSTM trained on syntactically impossible languages (reversed sentences, parity-based negations). GPT-2 shows lower perplexity on natural language (loss ratios up to 2.25× on reversed conditions), LSTM minimal differences. Authors propose functionalist paradigm against Chomsky's rationalist perspective.
Qualitative study with 22 creative writers on intentional AI refusals in writing assistance. Researchers explore how refusals (saying "no") could introduce reflective friction rather than blind compliance, depending on context (planning, drafting, reviewing) and individual preferences.
Study of verify-gated completion pattern for controlling persistent multi-agent systems. Bounded implementation: 99.5% verification success rate (1,791/1,800 events), 98.58% rule agreement with governance verifier. Results limited to decision inspectability and fail-closed behavior; no safety guarantees or task-level coverage claims supported.
FOL2NS is a neurosymbolic framework converting first-order logic formulas into natural language sentences. It handles deeply nested structures with varying quantifier depths rarely captured by existing corpora. Results show fluent generation but increasing challenges with structural complexity.
Survey on universal time-series representation learning. Proposes taxonomy based on three fundamental elements for state-of-the-art deep learning methods. Covers extraction of hidden patterns without manual feature engineering, with resources and future research directions.
VLMs struggle with planning from complex visual inputs. This paper proposes Pattern Induction, an online inductive learning strategy that discovers and optimizes reusable visual patterns as composite experts. Pattern Inference enables VLMs to recognize these patterns and directly infer world model structures. Evaluated on FrozenLake, Crafter, and CubeBench.
New paradigm for experience-driven learning in multimodal settings: agents learn to dynamically construct and organize memory based on task requirements and interaction history, rather than relying on fixed memory schemas. Adaptive memory design improves performance and generalization across multimodal tasks.
Theoretical paper on consent degradation in delegation chains between autonomous robots. Introduces CoRVE framework to verify consent across multi-agent architectures. Analyzes regulatory gaps in EU AI Act, GDPR, Machinery Regulation, and Product Liability Directive.
IsalProgram is a regular assembly-like language where every finite string is a valid program. Executed on a virtual machine with circular doubly linked list, it eliminates memory addresses and variable names. Authors prove its regularity and explore its potential for neural program synthesis.
Comprehensive review of AI methods for solving inverse partial differential equation (PDE) problems. Covers three categories: inverse problems, inverse design, and control. Applications: medical imaging, geophysics, aerodynamics, thermal systems. Challenges: physics-informed architectures, limited real-world data, uncertainty quantification, inverse foundation models.
Decentralized task allocation framework for circular manufacturing using Edge-AI and ranking-aware learning. Each machine evaluates tasks using local information (processing capability, queue state, resource contention). Results: reduced delays, improved deadline adherence, enhanced energy efficiency in discrete-event simulation.
Multidisciplinary framework for human oversight of AI systems in high-risk decision-making scenarios. Defines oversight architectures, actor roles, and implementation processes. Synthesizes open research challenges in this emerging field.
Paper introduces a reversible transformation between belief functions and possibilistic structures to overcome Dempster's rule limitations. Proposes a belief evolution network and triangular norm family for flexible evidential information fusion from non-distinct sources with improved conflict management.
Mixed-methods multi-course study at a New England university examining LLM impact on teaching and learning. Retrospective quantitative analysis, instructor and student surveys, historical grade data pre/post-LLM. Documents shifts in study methods, course design, and learning outcomes.
A 2023 Nature survey of 1,600 researchers shows tension between excitement and concern about AI tools in research. The article argues replacement is underway: shift from researcher-as-creator to researcher-as-curator. Key risk: humans retain responsibility while losing intellectual ownership and deep understanding of science.
Shared Backbone PPO algorithm for multi-UAV swarm communication coverage optimization. Sharing base module between Actor and Critic networks improves training efficiency. Graph information aggregation module integrated to model inter-agent communication conditions.
Research paper proposing to integrate AI advances into cognitive science methods to build generalizable theories. Authors argue that naturalistic experimental paradigms and computational models capable of accommodating them are necessary to understand natural intelligence and ensure theory generalization.
Sieve scans Cursor and Claude chat history to detect leaked API keys. Useful tool for identifying exposed secrets in conversations with AI assistants.
Vercel launches Flat Rate CDN in Limited Beta for Pro teams. This service replaces usage-based pricing with a fixed monthly fee, covering traffic spikes without overage charges.
Lemonade v10.5.1 released with MTP and ROCm 7.13 support for Strix Halo. Enables loading Qwen3.6-27B-MTP-GGUF with auto-applied MTP arguments. Also fixes Fedora 43 support.
Vercel enables monorepos to consolidate GitHub commit statuses into a single status per PR instead of one per project. Teams configure GitHub branch protection once and manage which Vercel projects are required for merge in each project's settings.
Vercel waives CDN Requests and Fast Data Transfer charges for traffic denied, challenged, or rate-limited by its Web Application Firewall (WAF). The change applies automatically to all projects using Vercel Firewall with no configuration needed.
NHS closes public GitHub repositories due to AI and security concerns. The decision aims to prevent AI model training on sensitive medical code and reduce security risks.
PaddleOCR 3.5 integrates a Transformers backend for OCR and document parsing tasks. The new version improves accuracy and flexibility by leveraging Transformers models, enabling better text recognition and structured data extraction.