The current paradigm of evaluating Large Language Models (LLMs) through static benchmarks comes with significant limitations, such as vulnerability to data contamination and a lack of adaptability to the evolving capabilities of LLMs.

The current paradigm of evaluating Large Language Models (LLMs) through static benchmarks comes with significant limitations, such as vulnerability to data contamination and a lack of adaptability to the evolving capabilities of LLMs. Therefore, evaluation methods that can adapt and generate evaluation data with controlled complexity are urgently needed. In this work, we introduce Dynamic Evaluation of LLMs via Adaptive Reasoning Graph Evolvement (DARG) to dynamically extend current benchmarks with controlled complexity and diversity. Specifically, we first extract the reasoning graphs of data points in current benchmarks and then perturb the reasoning graphs to generate novel testing data. Such newly generated test samples can have different levels of complexity while maintaining linguistic diversity similar to the original benchmarks.

Training large language models with Reinforcement Learning with Verifiable Rewards (RLVR) exhibits a set of distinctive and puzzling behaviors that remain poorly understood, including a two-stage learning curve, a V-shaped response-length trajectory, and a pronounced vulnerability to catastrophic forgetting. In this work, we propose that these behaviors are emergent collective phenomena governed not by neural implementation details, but by the topological evolution of the latent reasoning graph in semantic space. By demonstrating a dynamical isomorphism between a 1.5B-parameter LLM and a minimal Concept Network Model (CoNet), we trace the causal source to the self-organization of a sparse concept web pinned to an average degree of two. This geometric perspective provides a unified physical explanation for the observed anomalies: the V-shaped trajectory tracks the evolution from parallel local skill optimization to global network integration; catastrophic forgetting stems from the topological disconnection of critical ``trunk'' edges; and policy collapse arises from the accumulation of sequential transitions at the web's leaf nodes, where broad exploration abruptly freezes into rigid, high-reward trajectories. Identifying a ``maximally frustrated state'' at the transition between learning stages, we propose Annealed-RLVR, a principled algorithm that injects a targeted SFT ``heating'' step to resolve this topological bottleneck. Experiments confirm that this theory-driven intervention outperforms standard RLVR on both in-distribution and out-of-distribution benchmarks (including Minerva and AIME). By recasting RLVR from black-box optimization into a predictable process of structural self-organization, our work provides a new physical intuition for engineering the emergent reasoning capabilities of future AI systems.

Modern large language model-based reasoning systems frequently recompute similar reasoning steps across tasks, wasting computational resources, inflating inference latency, and limiting reproducibility. These inefficiencies underscore the need for persistent reasoning mechanisms that can recall and reuse prior computational traces. We introduce Graph-Memoized Reasoning, a formal framework for representing, storing, and reusing reasoning workflows as graph-structured memory. By encoding past decision graphs and retrieving them through structural and semantic similarity, our approach enables compositional reuse of subgraphs across new reasoning tasks. We formulate an optimization objective that minimizes total reasoning cost regularized by inconsistency between stored and generated workflows, providing a theoretical foundation for efficiency-consistency trade-offs in intelligent systems. We outline a conceptual evaluation protocol aligned with the proposed optimization objective. This framework establishes the groundwork for interpretable, cost-efficient, and self-improving reasoning architectures, offering a step toward persistent memory in large-scale agentic systems.

AI Scientific Assistant Core (AISAC) is an integrated multi-agent system developed at Argonne National Laboratory for scientific and engineering workflows. AISAC builds on established technologies - LangGraph for orchestration, FAISS for vector search, and SQLite for persistence - and integrates them into a unified system prototype focused on transparency, provenance tracking, and scientific adaptability. The system implements a Router-Planner-Coordinator workflow and an optional Evaluator role, using prompt-engineered agents coordinated via LangGraph's StateGraph and supported by helper agents such as a Researcher. Each role is defined through custom system prompts that enforce structured JSON outputs. A hybrid memory approach (FAISS + SQLite) enables both semantic retrieval and structured conversation history. An incremental indexing strategy based on file hashing minimizes redundant re-embedding when scientific corpora evolve. A configuration-driven project bootstrap layer allows research teams to customize tools, prompts, and data sources without modifying core code. All agent decisions, tool invocations, and retrievals are logged and visualized through a custom Gradio interface, providing step-by-step transparency for each reasoning episode. The authors have applied AISAC to multiple research areas at Argonne, including specialized deployments for waste-to-products research and energy process safety, as well as general-purpose scientific assistance, demonstrating its cross-domain applicability.

This report extends the Spectral Neuro-Symbolic Reasoning (Spectral NSR) framework by introducing three semantically grounded enhancements: (1) transformer-based node merging using contextual embeddings (e.g., Sentence-BERT, SimCSE) to reduce redundancy, (2) sentence-level entailment validation with pretrained NLI classifiers (e.g., RoBERTa, DeBERTa) to improve edge quality, and (3) alignment with external knowledge graphs (e.g., ConceptNet, Wikidata) to augment missing context. These modifications enhance graph fidelity while preserving the core spectral reasoning pipeline. Experimental results on ProofWriter, EntailmentBank, and CLUTRR benchmarks show consistent accuracy gains (up to +3.8\%), improved generalization to adversarial cases, and reduced inference noise. The novelty lies in performing semantic and symbolic refinement entirely upstream of the spectral inference stage, enabling efficient, interpretable, and scalable reasoning without relying on quadratic attention mechanisms. In summary, this work extends the Spectral NSR framework with modular, semantically grounded preprocessing steps that improve graph quality without altering the core spectral reasoning engine. The result is a more robust, interpretable, and scalable reasoning system suitable for deployment in open-domain and real-world settings.

The current paradigm of evaluating Large Language Models (LLMs) through static benchmarks comes with significant limitations, such as vulnerability to data contamination and a lack of adaptability to the evolving capabilities of LLMs.

However, there is an ongoing debate about the extent to which LLMs are capable of actual reasoning .

LLM-generated reasoning graphs, referred to as mission-specific graphs (MSGs), are increasingly used for video anomaly detection (VAD) and recognition (VAR). These MSGs are novel artifacts: they often exhibit skewed connectivity and lack large-scale datasets for pre-training, which makes existing graph structure refinement (GSR) methods ineffective. To address this challenge, we propose HDC-constrained Graph Structure Refinement (HDC-GSR), a paradigm that leverages hyperdimensional computing (HDC) to optimize decodable graph representations without relying on structural-distribution learning. Building on this paradigm, we introduce MissionHD, an HDC framework that encodes graphs with constrained graph-neural operations, aligns them directly with downstream task loss, and decodes refined structures. Experiments on VAD/VAR benchmarks demonstrate that MissionHD-refined graphs consistently improve performance, establishing HDC-GSR as an effective pre-processing step for structured reasoning in video anomaly tasks.