Tutorial on multilingual multimodal LLMs for low-resource languages. Covers recent models (PALO, Maya), speech-text-vision pipelines, low-cost data creation, tri-modal alignment via adapters, and culture-aware evaluation beyond English.
Paper proposing an experiment to establish case law on inventorship in AI-assisted inventions. Given the lack of legal precedent and rapid AI development, authors suggest a structured approach involving stakeholders to identify how to measure human contribution and determine inventor recognition.
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 arXiv paper proposing GLGAT (Global-Local Graph Attention Network) for traffic forecasting. The model combines a global attention matrix for the entire graph with local attention matrices per vertex, using pairwise encoding and event-based adjacency matrix. Experiments on two real-world traffic datasets show competitive performance against state-of-the-art baselines.
Thesis proposing a modular framework for reasoning over uncertainty in knowledge graphs at three levels: imprecise attribute values, probabilistic triple existence, and incomplete schema. Combines probabilistic literals, tractable probabilistic circuits via SPARQL compilation, and topology-aware geometric embeddings for schema reasoning.
Review of AI trends in cybersecurity focused on intrusion detection. Comparative analysis of approaches using generative AI, NLP, federated learning, and explainable AI to improve interpretability and trust in detection systems.
Data-driven approach to integrate constructs and their relations in information systems. Combines task-adapted text embeddings and clustering to group constructs from structural equation models. Optimizes trade-off between semantic purity and parsimony through explicit loss function.
Decentralized building management framework integrating DAOs, digital twins, LLMs, and blockchain for transparent governance. System evaluated on cost efficiency, scalability, data security, and usability via System Usability Scale and expert interviews.
Agentic AI Translate is an agentic translator prototype replacing the text-in/text-out paradigm with a four-stage cycle (Identify → Prompt → Generate → Verify). Users compose a structured translation brief through model-assisted dialogue grounded in skopos theory, register, and genre conventions. Verification uses the GEMBA-MQM error-span protocol for evidence-grounded scoring.
Theoretical paper proposing a unified framework for phase transitions in deep learning (grokking, emergent capabilities) and non-equilibrium chemistry. Introduces two gradient fields (entropy production rate and information quasi-potential) and two order parameters (adversarial breakdown threshold α†, self-referential coupling threshold κc) to describe driven informational systems.
Comparative study of BERT (encoder-only) and T5 (seq2seq) for named entity recognition (NER). BERT uses a classification head with weighted cross-entropy; T5 fine-tuned with few-shot prompts. Evaluation on 7-class and simplified 3-class tag schemes, including ablation study and error analysis.
arXiv study on automating Motivational Interviewing (MI) session coding for alcohol use reduction. Uses audio-language models (ALMs) with 4 complementary analytic prompts and self-consistency (12 reasoning trajectories per utterance). On 5 sessions, achieves 52.56% accuracy, 54.03% precision, 47.45% recall, macro-F1 46.40%, exceeding baselines.
Proposes quantum sidecar architectures for hybrid AI training and inference. Two operating modes: stateful protected-register mode (QND readout with ancilla) and stateless reset-and-reprepare mode (QAOA-style circuits). Simulations on 2/4/6/8 protected qubits. Positions quantum sidecars as bounded signal generators for optimization, expert selection, and routing.
Theoretical paper developing a proof-theoretic semantic account of information grounded in inferentialist reasoning. Replaces truth with inferability in Dretske's framework, introduces the 'inferon' as a primitive unit, and applies proof-theoretic tools to distributed systems modelling.
Safety-critical control method for autonomous navigation in unknown environments. Uses conformal prediction to generate risk-aware obstacle ellipsoids accounting for sensor uncertainty, then two nested differentiable optimization layers to build control barrier functions. Validated through numerical simulations.
Study of Dante's Divina Commedia using vowel-consonant encoding modeled as a four-state Markov chain. Graphemic memory index increases gradually from Inferno to Paradiso. Trigram analysis reveals recurrent configurations linked to lexical environments and orthographic phenomena.
Residual Semantic Decomposition (RSD) recursively decomposes word embeddings into local semantic axes via neural additive approach. On ambiguous words, RSD separates supplied contexts from shuffled controls, but entropy diagnostics show static GloVe does not uniformly place ambiguous words at high-entropy boundaries.
EGI is a multimodal framework to monitor unconscious emotions of Scrum Masters in real-time. The system combines speech-to-text transcription (WER 10%), prosody analysis, emotional vocabulary matching, and context-aware suggestions via open-source multi-module API. Testing shows significant improvement in emotional awareness during simulated agile meetings.
Position paper arguing that generalizable agents require environment scaling—expanding the distribution of executable rule-sets agents interact with—beyond trajectory or task scaling within fixed benchmarks. Proposes unified taxonomy separating trajectory, task, and environment scaling; synthesizes construction paradigms (programmatic generators vs generative world models) for scalable environments.
Analysis of 60k German news articles covering 5.5k landslide events over 25 years. Reveals overreporting of Southern and Western Europe relative to actual landslide susceptibility. Study of spatial bias in media coverage of international natural disasters.
PuppyChatter is a software framework designed to simplify development of LLM-based AI applications. It combines the simplicity of vendor-specific SDKs with vendor-neutrality principles of abstraction frameworks, reducing complexity and vendor lock-in risks.
EHV is a JIT compiler architecture embedding formal verification of AI governance policies directly into the inference pipeline. Using CRDTs and TEEs, it achieves sub-millisecond formal determinism (SMFD) and reduces governance latency from days to O(1), eliminating the trade-off between deployment velocity and compliance.
Position paper on challenges of modeling human haptic material perception with AI. Identifies three bottlenecks: scarcity of large, diverse haptic datasets; lack of standardized evaluation platforms and perceptual benchmarks; limitations in model capacity and interpretability. Calls for cross-disciplinary efforts to advance.
Perspective paper on using LLMs (ChatGPT, Gemini, Claude) and specialized tools (Consensus, Elicit) for conducting literature reviews. Covers summarization, question-answering, data extraction, and translation. Discusses opportunities, risks, and philosophical implications for scientific progress.
UbiSLAM presents a real-time mapping and localization system for dynamic indoor environments using a network of fixed RGB-D cameras. This approach reduces computational load on robots and improves navigation accuracy and human-robot interactions, while requiring automatic calibration and optimized communication protocols to handle blind spots.
Vision paper on AI-assisted code review. Proposes framework of specialized agents spanning 5 stages (PR creation, augmentation, reviewer selection, AI-assisted review, retrospective) with human control at critical gates. Identifies challenges: reliability, bias, privacy, automation bias, transparency, evaluation.
Experimental study with 300 UK participants measuring individual preferences for well-being and fairness. Utility functions estimated via Expected Utility Maximisation reveal inequality aversion unrelated to political alignment. Results challenge average life satisfaction as a policy metric and support nonlinear utility-based alternatives for value-aligned AI systems.
Theoretical paper proposing a unified framework for phase transitions in deep learning (grokking, emergent capabilities) and non-equilibrium chemistry. Introduces two gradient fields (entropy production rate and information quasi-potential) and two order parameters (adversarial breakdown threshold α†, self-referential coupling threshold κc) defining a candidate universality class.
WorldString is a neural architecture modeling the state manifold of real-world objects from point clouds or RGB-D video streams. Designed as a differentiable digital twin, it serves as a foundational building block for physical world models integrating policy learning and neural dynamics.
Theoretical paper arguing AI in biomedicine must transition from predictive systems based on historical data to interventional models capable of simulating novel treatment effects. Current architectures remain observational and cannot generalize to unobserved interventions.
Automatic generation of feedback fuzzy cognitive maps (FCMs) from text using LLM agents to chunk text with overlaps. Convex mixing of chunk FCMs produces representative cyclic FCM knowledge graphs. Operator-level Bayesian inference generates de-chunked posterior FCMs. Demonstrated on Allison's Thucydides Trap model: 7 out of 8 FCM knowledge graphs predicted war when stimulated.
Firefly Algorithm variant for automatic clustering. Introduces centroid movement strategy and multi-objective fitness function (compactness, separation, TSP-based penalty). Automatically estimates optimal cluster count. Outperforms K-Means on robotic sensor networks.
Method using homoglyph-based adversarial perturbation to modify computer science problem statements without changing semantic meaning. Aims to prevent ChatGPT, Gemini, and Claude from directly solving student homework. Interactive tool provided.
Study on Bangla fake news detection using CNN. Combines semantic, statistical, and character-level features on BanFakeNews-2.0 dataset. Results show that hybrid feature combinations significantly improve recall and F1-scores compared to individual features.
Theoretical paper on computational challenges of token economics in LLM systems. Introduces the "Token Economics Trilemma": tensions between fine-grained valuation, low-latency execution, and allocation optimality. Identifies three technical areas: real-time value accounting, constrained resource allocation, and economic-aware system architecture.
Position paper arguing that agent generalization requires environment scaling—expanding the distribution of executable rule-sets agents interact with, beyond trajectory or task scaling. Proposes unified taxonomy separating trajectory scaling, task scaling, and environment scaling. Contrasts programmatic generators with generative world models for constructing scalable environments.
LLMCap is a proxy that automatically stops LLM API calls when a dollar spending cap is reached. Cost control tool to prevent budget overruns when using LLM APIs.
Google and Blackstone form a joint venture for an AI cloud platform with in-house chips. The initiative aims to reduce vendor dependency and provide optimized AI infrastructure to institutional clients.
SuperInfer introduces rotary scheduling and memory management for LLM inference optimized to meet SLO (Service Level Objectives). System-level approach to reduce latency and memory consumption.
Simon Willison summarizes six months of LLM developments in a five-minute lightning talk at PyCon US 2026. The November 2025 inflection point marks a critical turning point, especially for coding. The best model changed hands 5 times between Anthropic, OpenAI, and Google.