Skip to content

References: Research Workflows and the Researcher as Orchestrator

  1. Generative artificial intelligence - Wikipedia - Surveys generative models that produce text, code, images, and audio, with history, applications, and limitations. Supports the chapter's contrast between generative and discriminative models and its warnings about plausible but wrong output.

  2. Large language model - Wikipedia - Explains how LLMs are trained to predict tokens, their architecture, and their known limits such as hallucination. Background for the chapter's next-token illustration, knowledge cutoff, and context window discussion.

  3. Intelligent agent - Wikipedia - Defines agents that perceive an environment and act toward goals, and classifies them from simple reflex to learning agents. Grounds the chapter's goal, tools, loop, and memory definition of an AI agent.

  4. Artificial Intelligence: A Modern Approach (4th Edition) - Stuart Russell and Peter Norvig - Pearson - Organizes the whole field around the rational agent, using the percept, action, and environment framing and the PEAS task-environment description. This clarifies how an agent differs from a one-shot assistant.

  5. The Craft of Research (4th Edition) - Wayne C. Booth, Gregory G. Colomb, Joseph M. Williams, and Joseph Bizup - University of Chicago Press - Builds a question from topic, question, and significance. Its "so what?" framing sharpens vague interests into the specific, falsifiable research questions this chapter requires.

  6. Building effective agents - Anthropic - Distinguishes predefined workflows from agents that direct their own tool use, and catalogs patterns such as prompt chaining and orchestrator-workers. Directly supports the chapter's task decomposition and orchestration ideas.

  7. ReAct: Synergizing Reasoning and Acting in Language Models - arXiv - Introduces interleaving reasoning steps with actions so a model can plan, consult sources, and adjust. A primary-source account of the decide, act, observe loop that defines AI agents.

  8. Good Enough Practices in Scientific Computing - arXiv - Practical recommendations on data management, software, project organization, and version control for researchers. Supports the chapter's reproducibility discipline of named artifacts, recorded inputs, and scripted analysis.

  9. CS324: Large Language Models (Winter 2022) - Stanford University - Course notes covering how language models are built, their capabilities, harms, and societal impact. Useful for readers who want depth beyond the chapter's vocabulary of LLMs, hallucination, and limits.

  10. Claude Code overview - Anthropic - Official documentation for an agentic coding tool that reads a codebase, edits files, runs commands, and supports subagents and instruction files. A concrete example of the orchestrator's workflow.