arXiv subjects

Pouya Ghiasnezhad Omran

Publications and source records attributed to Pouya Ghiasnezhad Omran.

4 recordsLinked to original sources

arXiv2026

Agent Gym: A Framework for Continuous Evaluation and Evolution of LLM Agents Through Human-in-the-Loop Feedback

Large Language Model (LLM) agents deployed in production environments face a fundamental tension: the agent's behavior is frozen at deployment time, while the business rules and edge cases it must handle continue to evolve. Existing approaches address agent construction and one-time evaluation but provide no structured mechanism for continuous post-deployment behavioral correction without modifying the agent's source code. Most of the approaches offered in the market, require intense collection of logs and traces, and re-examining the agent design by the engineering team, a process which is heavy, long and negates the economical value of agentic transformation. We introduce Agent Gym, a modular, domain-agnostic framework that wraps any existing LLM-based agent in a continuous evaluation-and-evolution loop. The framework provides six composable capabilities --- Act, Evaluate, Investigate, Correct, Learn, and Observe --- organized across three architectural zones: a constitution layer that codifies domain knowledge in configuration artifacts, a runtime inference pipeline that chains acting, investigation, and adaptive correction, and a learning loop that enables subject matter experts to discover and validate new correction rules through natural language interaction. The key technical contributions include a hybrid deterministic-LLM correction engine with 21 condition operators and three-tier actions, a three-layer investigation architecture for ground-truth-free compliance validation, and a programmatic safety loop that guarantees rule correctness before human approval. We further introduce the Spec-to-Note Gap, an autoencoder-inspired view of agentic system transparency. An open-source reference implementation for invoice processing demonstrates that the framework is fully operational and ready for adoption.

cs.AI↗

arXiv2026

Muscle Memory for Agents: Compile not Merely Retrieve

Memory for LLM agents has converged on a single architectural pattern: store experience as text, embeddings, reflections, or rules; retrieve at inference time; let a general-purpose orchestrator interpret what to do. This paper argues that the pattern is the wrong default for personalization. We position Muscle Memory - the practice of compiling recurring user intent into purpose-built specialist agents - as a distinct memory paradigm from retrieval, and we argue that compilation is a better fit for the workloads where current assistants impose a multi-turn tax on their users: making them repeatedly correct format, depth, and scope to obtain a domain-appropriate answer. We support the position with a reference implementation and empirical evidence. The implementation is a four-phase pipeline (Harvest $\rightarrow$ Analyze $\rightarrow$ Augment $\rightarrow$ Evaluate) that mines conversational history, separates behavioral from task patterns, and emits quality-gated executable compiled specialists with two-stage trigger matching. On 90 held-out scenarios across five user personas, the augmented assistant wins 32 of 36 cases where a specialist fires, an 88.9% win rate, with a +2.05 personalization gain and only a $-0.28$ accuracy cost on a 1-4 scale. We discuss why compilation is better suited than retrieval in this regime, what the result implies for the broader memory design space, and what open problems remain.

cs.MA↗

arXiv2026

A Systematic Comparison and Evaluation of Building Ontologies for Deploying Data-Driven Analytics in Smart Buildings

Ontologies play a critical role in data exchange, information integration, and knowledge sharing across diverse smart building applications. Yet, semantic differences between the prevailing building ontologies hamper their purpose of bringing data interoperability and restrict the ability to reuse building ontologies in real-world applications. In this paper, we propose and adopt a framework to conduct a systematic comparison and evaluation of four popular building ontologies (Brick Schema, RealEstateCore, Project Haystack and Google's Digital Buildings) from both axiomatic design and assertions in a use case, namely the Terminological Box (TBox) evaluation and the Assertion Box (ABox) evaluation. In the TBox evaluation, we use the SQuaRE-based Ontology Quality Evaluation (OQuaRE) Framework and concede that Project Haystack and Brick Schema are more compact with respect to the ontology axiomatic design. In the ABox evaluation, we apply an empirical study with sample building data that suggests that Brick Schema and RealEstateCore have greater completeness and expressiveness in capturing the main concepts and relations within the building domain. The results implicitly indicate that there is no universal building ontology for integrating Linked Building Data (LBD). We discuss ontology compatibility and investigate building ontology design patterns (ODPs) to support ontology matching, alignment, and harmonisation.

cs.IR↗

arXiv2023

A System's Approach Taxonomy for User-Centred XAI: A Survey

Recent advancements in AI have coincided with ever-increasing efforts in the research community to investigate, classify and evaluate various methods aimed at making AI models explainable. However, most of existing attempts present a method-centric view of eXplainable AI (XAI) which is typically meaningful only for domain experts. There is an apparent lack of a robust qualitative and quantitative performance framework that evaluates the suitability of explanations for different types of users. We survey relevant efforts, and then, propose a unified, inclusive and user-centred taxonomy for XAI based on the principles of General System's Theory, which serves us as a basis for evaluating the appropriateness of XAI approaches for all user types, including both developers and end users.

cs.AI↗