Theme: llm-reasoning

agentic-ai knowledge-graphs memory-context llm-reasoning ai-architecture

How can Artificial Intelligence (AI) agents effectively synchronize structured semantic memory, meaning ontologies and knowledge graphs, with latent knowledge encoded in Large Language Model (LLM) wei…

Episodic-to-semantic memory consolidation in AI agents

agentic-ai memory-context knowledge-graphs llm-reasoning ai-architecture

What techniques enable AI agents to reliably generalize from specific episodic experiences (interaction logs, task traces, observed events) to durable semantic memory entries (ontological facts, proce…

Autonomous knowledge curation and truth maintenance for agentic ontologies

agentic-ai knowledge-graphs knowledge-management llm-reasoning governance-policy

What mechanisms exist, or are under active research, to enable Artificial Intelligence (AI) agents to autonomously curate which extracted knowledge is worth retaining in a long-term ontology, detect a…

TBox-driven vs ABox-emergent ontology approaches in GraphRAG systems

knowledge-graphs rag-retrieval benchmarks-eval llm-reasoning

To what extent do TBox (Terminological Box)-driven (predefined upper- and mid-level) ontologies outperform, underperform, or complement ABox (Assertion Box)-emergent (bottom-up, data-driven) approache…

Ontology Completeness as a World Model for Large Language Model (LLM) Prediction

2026-05-25

llm-reasoning knowledge-graphs agentic-ai ai-architecture

To what extent can a sufficiently complete ontology function as a practical world model (in the sense described by Yann LeCun) for Large Language Models (LLMs) making predictive inferences, and which…

LLM reasoning in mathematics and programming tasks

2026-05-25

llm-reasoning formal-methods benchmarks-eval software-engineering ai-architecture

To what extent is the claim true that mathematics and programming are especially strong use cases for Large Language Models (LLMs) because both rely on formal symbolic languages that may align with mo…

Theory and mechanisms of prompt and program optimization in Language Models

2026-05-21

llm-reasoning benchmarks-eval

What theory best explains why prompt and program optimization methods can outperform baseline prompting and Reinforcement Learning (RL) in Language Model (LM) pipelines, and how do the methods in the…

Stochastic LLM Agent vs. Deterministic Coded System

agentic-ai llm-reasoning security-risk ai-architecture formal-reliability

How do the failure modes of a stochastic multi-step Large Language Model (LLM) agent, meaning a tool-using system whose action path can vary across runs, differ fundamentally from the failure modes of…

Formal Generalisation Bounds for Tool-Using LLM Systems When Tools Return…

agentic-ai llm-reasoning tools-infrastructure formal-guarantees distribution-shift

What formal bounds can be stated for generalisation outside the training distribution in tool-using Large Language Model systems when their tools return non-deterministic outputs under unconstrained p…

Agentic Tool-Feedback Loops and Explanatory Reach

agentic-ai llm-reasoning benchmarks-eval ai-architecture

When a Large Language Model (LLM) is wrapped in an agentic loop, meaning a repeated perception, strategy-selection, tool-action, and verification cycle, does the outer loop introduce true explanatory…

In-Context Learning and Chain-of-Thought Prompting

llm-reasoning benchmarks-eval consciousness-cognition causal-inference

What are the empirical boundaries of in-context learning and chain-of-thought prompting when they are used to push a purely predictive statistical architecture toward intervention questions and altern…

The Stochastic Parrot Under Pressure

llm-reasoning benchmarks-eval consciousness-cognition causal-inference

How does the Stochastic Parrot hypothesis, the claim that Large Language Models (LLMs) reproduce linguistic form more readily than grounded structural understanding, manifest when an LLM is presented…

Large Language Models as Statistical Optimisers

llm-reasoning benchmarks-eval ai-architecture causal-representation-learning epistemology-of-ai

To what extent do Large Language Models (LLMs) optimise strictly for linguistic form and statistical token distribution rather than constructing internal, invariant causal models of reality?

Pearl's Causal Hierarchysynthesis

llm-reasoning ai-architecture causal-reasoning formal-epistemology

What are the formal information-theoretic boundaries that prevent a model trained exclusively on observational data (Level 1 on Pearl's Ladder of Causation) from ever executing or predicting the outco…

David Deutsch's Hard-to-Vary Criterion

llm-reasoning ai-architecture benchmarks-eval epistemic-formalism model-interpretability formal-methods

Using David Deutsch's hard-to-vary criterion, meaning an explanation whose details cannot be changed without losing explanatory force, what formal criteria can measure the internal logical constraints…

LLM-First Policy Clarification and Institutional Knowledge Atrophy

llm-reasoning governance-policy workforce-skills knowledge-management organisational-learning agentic-ai organisational-design tools-infrastructure

How does shifting from peer policy clarification to Large Language Model (LLM)-first interaction affect institutional memory transfer, mentoring, and long-term policy expertise?

Policy Quality Degradation and Cross-Institution Blind Spots When New Policy…

llm-reasoning governance-policy security-risk benchmarks-eval organisational-impact agentic-ai organisational-design tools-infrastructure

What policy-quality degradation and systemic blind-spot risks emerge when organisations draft new policy versions from Large Language Model (LLM) interpretations of previous policy versions?

LLM Training Prior Contamination in Compliance Interpretation

llm-reasoning memory-context governance-policy security-risk agentic-ai organisational-design benchmarks-eval

What failure modes emerge when Large Language Models (LLMs) combine generic public legal knowledge with proprietary organisational policy in compliance interpretation tasks?

Adversarial prompting risks in policy assistants

agentic-ai governance-policy security-risk llm-reasoning organisational-design benchmarks-eval

How vulnerable are corporate compliance Large Language Models (LLMs) to adversarial prompting that reframes restrictive policy as permissive guidance, and which controls detect or contain deliberate m…

AI-Assisted Policy Interpretation and Accountability Displacement

governance-policy security-risk llm-reasoning organisational-design tools-infrastructure agentic-ai

How does integration of Large Language Models (LLMs) into policy-ambiguity resolution change liability allocation, escalation behaviour, and an organisation's ability to justify the resulting decision…

Layered reasoning stack interfaces

agentic-ai llm-reasoning ai-architecture governance-policy formal-methods tools-infrastructure

What state abstraction boundaries and interface protocols are most effective for mapping Large Language Model (LLM) candidate outputs into Energy-Based Model (EBM) evaluation state spaces while preser…

Agent-to-Agent (A2A)-to-tool-calling unification

2026-05-13

agentic-ai multi-agent tools-infrastructure llm-reasoning ai-architecture

To what extent does unifying specialised Agent-to-Agent (A2A) protocols into a standardised tool-calling interface affect orchestration overhead and reasoning accuracy in hierarchical multi-agent syst…

Security, Compliance, and Governance Risks of Using Generative AI (GenAI) Tools…

2026-05-10

governance-policy security-risk agentic-ai llm-reasoning

What are the documented security, compliance, and governance risks of using Generative Artificial Intelligence (GenAI) tools such as Microsoft 365 (M365) Copilot for drafting memos, reports, and other…

Practical Limits of Large Language Model (LLM) Determinism

2026-05-09

llm-reasoning benchmarks-eval cost-performance governance-policy agentic-ai

What are the practical limits of making LLM (Large Language Model)-based decisions or policy enforcement deterministic, even with temperature=0, fixed seeds, and constrained prompts?

Hybrid Architecture Design

2026-05-09

agentic-ai governance-policy ai-architecture llm-reasoning tools-infrastructure

How should hybrid architectures be designed so that probabilistic LLMs handle interpretation and insight generation while deterministic layers enforce final governance, compliance, and high-stakes dec…

Compliance Risks of Relying on Stochastic Large Language Model (LLM) Outputs…

2026-05-09

governance-policy security-risk llm-reasoning agentic-ai

What evidence or guidance exists on the compliance risks of relying primarily on stochastic Large Language Model (LLM) outputs for governance, privacy, or regulatory decisions?

How do open-weight policy enforcement reasoning models, exemplified by OpenAI's…

llm-reasoning governance-policy security-risk mlops-deployment cost-performance benchmarks-eval tools-infrastructure agentic-ai

How do open-weight, meaning released-weight and self-hostable, policy enforcement reasoning models, exemplified by OpenAI's gpt-oss-safeguard, classify text against strict, customizable policies, and…

How does Factual precision Scoring (FActScore) operationalise atomic-level…

benchmarks-eval llm-reasoning

How does FActScore (Factual precision Scoring), developed at the University of Washington, operationalise the concept of atomic factual claim decomposition and precision scoring for Large Language Mod…

How can findings from OpenFactCheck, Loki, FActScore, gpt-oss-safeguard, and…synthesis

benchmarks-eval llm-reasoning tools-infrastructure

How can the findings from research into OpenFactCheck, Loki, FActScore, gpt-oss-safeguard, and Barnum statement identification techniques be synthesised into concrete, actionable improvements to the a…

What are Barnum statements (Forer Effect statements), how do they manifest in…