AI
-
AgentArch: A Comprehensive Benchmark to Evaluate Agent Architectures in Enterprise
Source: arXiv AI PapersRead more: AgentArch: A Comprehensive Benchmark to Evaluate Agent Architectures in EnterpriseThis study offers a comprehensive benchmark for evaluating 18 distinct agentic configurations within advanced large language models. Key dimensions assessed…
-
LLM Enhancement with Domain Expert Mental Model to Reduce LLM Hallucination with Causal Prompt Engineering
Source: arXiv AI PapersRead more: LLM Enhancement with Domain Expert Mental Model to Reduce LLM Hallucination with Causal Prompt EngineeringThis paper discusses the application of large language models (LLMs) and Retrieval-Augmented Generation (RAG) for improving decision-making processes. It addresses…
-
Is the `Agent’ Paradigm a Limiting Framework for Next-Generation Intelligent Systems?
Source: arXiv AI PapersRead more: Is the `Agent’ Paradigm a Limiting Framework for Next-Generation Intelligent Systems?This paper critiques the agent-centric paradigm in AI, discussing its limitations and biases. It suggests a shift towards system-level dynamics…
-
Enhancing Computational Cognitive Architectures with LLMs: A Case Study
Source: arXiv AI PapersRead more: Enhancing Computational Cognitive Architectures with LLMs: A Case StudyThis article discusses the integration of large language models (LLMs) with computational cognitive architectures, specifically the Clarion architecture. The research…
-
Rethinking Human Preference Evaluation of LLM Rationales
Source: arXiv AI PapersRead more: Rethinking Human Preference Evaluation of LLM RationalesThis study investigates the evaluation of rationales generated by large language models (LLMs). It proposes an attribute-based approach for assessing…
-
Free-MAD: Consensus-Free Multi-Agent Debate
Source: arXiv AI PapersRead more: Free-MAD: Consensus-Free Multi-Agent DebateFree-MAD is a new framework designed to enhance the reasoning abilities of large language models by eliminating the need for…
-
Tractable Asymmetric Verification for Large Language Models via Deterministic Replicability
Source: arXiv AI PapersRead more: Tractable Asymmetric Verification for Large Language Models via Deterministic ReplicabilityThis paper presents a novel verification framework to address trust issues in multi-agent systems utilizing Large Language Models (LLMs). It…
-
Difficulty-Aware Agent Orchestration in LLM-Powered Workflows
Source: arXiv AI PapersRead more: Difficulty-Aware Agent Orchestration in LLM-Powered WorkflowsA new framework, Difficulty-Aware Agentic Orchestration (DAAO), aims to enhance the efficiency of multi-agent systems using Large Language Models (LLMs).…
-
Neural cellular automata: applications to biology and beyond classical AI
Source: arXiv AI PapersRead more: Neural cellular automata: applications to biology and beyond classical AINeural Cellular Automata (NCA) combine traditional rule-based systems with trainable neural networks to model biological self-organization. This framework has applications…
-
AlignKT: Explicitly Modeling Knowledge State for Knowledge Tracing with Ideal State Alignment
Source: arXiv AI PapersRead more: AlignKT: Explicitly Modeling Knowledge State for Knowledge Tracing with Ideal State AlignmentAlignKT proposes an innovative approach to knowledge tracing in intelligent tutoring systems by explicitly modeling a stable knowledge state. It…