large language models

Preparing Citizens for Emergency Calls with a Hybrid FSM-LLM Dialogue Agent

Emergency calls are time-critical, verbal-only interactions in which call takers must assess severity and make decisions based solely on the caller’s description. Effective communication is critical during these calls, especially since most callers are inexperienced and untrained due to the rarity of emergency situations. To address this challenge, we develop a task-oriented dialogue agent that simulates emergency call takers to prepare citizens for effective medical emergency communication. It uses a finite-state machine as its dialogue policy and large language models for natural language understanding. This agent conducts protocol-guided interactions and maps caller descriptions to structured symptoms. In a small exploratory study with human participants, our agent achieved higher observed classification accuracy and dialogue efficiency while maintaining high naturalness scores.

Evolving transition-state search with agentic large language models

Abstract Accurate identification of transition states (TSs) is fundamental to computational chemistry. Modern reaction-discovery efforts increasingly rely on curating and completing large reaction datasets, where even a small fraction of TS-search failures can leave key pathways unresolved and bias the resulting reaction network. Here, we introduce an adaptive evolutionary framework for TS search workflows that prioritizes completion of a target dataset over global algorithmic robustness. The framework iteratively focuses on reactions for which no validated TS has been found and dynamically modifies the workflow to search the remaining TSs. Rather than converging toward a single universally optimal TS search workflow, the framework generates a sequence of specialized workflow variants that collectively increase TS coverage across the dataset. Applied to the Transition1X benchmark (over 10,000 reactions), this adaptive evolution improves a state-of-the-art TS-search workflow to achieve a >80% success rate of TS-finding with rigorous validation via eigenvector analysis and reaction path endpoint confirmation. More broadly, these results suggest that adaptive, LLM-driven evolutionary workflow optimization provides a transferable strategy for improving validated TS coverage in large-scale, failure-prone scientific workflows.

Behavioral collapse to longest-processing-time scheduling in contract-validated single-prompt large language model heuristic synthesis for identical parallel machine scheduling

Educators’ Discussion with a Generative Artificial Intelligence Trainer

It is learned that genAI was created to be people pleasing which can override its ability to find truth, and how it supports error analysis in Large Language Models, and how the public should view these tools.

Harmonized recommendation with large language model guidance and reinforcement learning

A dual-layer bilevel optimization model to complete algorithm evolution in agile satellite task scheduling problem using Large Language Models

Prompt engineering for medical named entity recognition: A few-shot approach with large language models

Do large language models displace or complement investment advisors? Evidence from Chinese wealth management firms

Data and Supplementary Materials for "Quantifying Materiality Misalignment Risk in ESG Reporting Using Large Language Models"

Canonical Noncommutative Defects Beyond Parikh Projection: Magnus Towers and an Unbounded Static-Dynamic Boundary Gap

We develop an order-sensitive defect formalism for substitution prefixes using the classical Magnus expansion of words. A substitution morphism induces a filtered endomorphism of the completed noncommutative tensor algebra, and an anchored substitution cut produces a multiplicative residual whose finite truncations form a compatible tower. The degree-one truncation recovers Parikh information, while higher degrees record scattered-subword data. For a finite substitution boundary language, we define the static separation depth $K_*$, the first Magnus depth at which every boundary word is distinguished, and for constant-length substitutions we define a dynamic stabilization depth $K_{\mathrm{dyn}}$ by minimizing the corresponding depth-$k$ boundary-output automata. We prove: $$K_{\mathrm{dyn}} \le K_*$$ and give exact depth-two and depth-three models. The main result is an explicit primitive binary constant-length family $\sigma_K$ such that: $$K_*(P_{\sigma_K}) = K \quad \text{but} \quad K_{\mathrm{dyn}}(\sigma_K) = 1 \quad \text{for every } K \ge 2$$ Hence, the static order depth and dynamic defect depth can differ by an arbitrarily large amount. The proof explicitly separates classical input—Magnus expansions, $k$-binomial equivalence, Thue–Morse separation results, and automaton minimization—from the canonical substitution-boundary architecture and the resulting static-dynamic separation theorem.

Consciousness Is a Phase Transition: A Two-Channel Thermodynamic Criterion and an Exclusion Principle

We propose that consciousness is a phase transition of a non-equilibrium dissipative structure, characterised by two independently controlled channels that must both be driven past threshold. The central result is an exclusion criterion: a macroscopic dissipative structure cannot be conscious unless it satisfies three necessary conditions - (i) an energy channel G > 1 maintaining non-equilibrium pumping, (ii) an information channel G_info(kappa) > 1 maintaining global phase coherence (with analytic critical point kappa_c = 1.8809), and (iii) topological closure via a self-referential triad (|SCC| >= 3). The framework traces a single causal chain: non-equilibrium driving -> Brusselator Hopf bifurcation -> sigmoid threshold -> phase-amplitude coupling (PAC) -> von Mises phase density -> Kuramoto network amplification -> G_info > 1 -> conscious phase transition. Each stage is derived from first principles (Appendices C-G). At the microscopic level, a basis-free trace-distance order parameter O(g) = 1/2 tr|rho+ - rho-| in two-qubit Gorini-Kossakowski-Sudarshan-Lindblad (GKSL) systems rigorously proves that the energy and information channels are independently controllable: a symmetrisation scan collapses O(g) from 0.306 to < 10^-4 as the absorption/emission asymmetry is removed, while a separate dephasing scan destroys coherence-carried asymmetry without affecting the energy flow. The topological closure condition is instantiated at multiple physical scales by a self-referential triad of three irreducible roles - reversible carrier, energy currency, irreversible anchor - which we identify in quantum coherence, aerobic metabolism, neural dynamics, and planetary geochemistry. We demonstrate the framework's explanatory power through clinical neuroscience and artificial intelligence. General anaesthesia abolishes consciousness by selectively collapsing kappa below threshold while metabolic pumping persists - a channel dissociation no single-channel theory predicts. Mindfulness meditation acts as a phase-locking mechanism that elevates kappa past kappa_c by suppressing Default Mode Network noise. Feedforward large language models fail all three conditions: their computational graph is acyclic, sustains no limit cycle, and operates as a closed system at inference. The widely observed model collapse under recursive self-training is the thermodynamic signature of the system relaxing toward equilibrium.

A Large Language Model Agent Framework for Token-Efficient Tabular Data Error Detection

Reproducibility materials for the PeerJ manuscript “A Large Language Model Agent Framework for Token-Efficient Tabular Data Error Detection”. This release contains the LAED source code, five aligned dirty/clean benchmark CSV pairs, the dependency specification, and documentation required to reproduce the supplied experiments. GitHub release: https://github.com/wangpy-gz/LAED/releases/tag/v1.0.2 Commit: e3b81e6