I do not think any simple modification of the concept of an agent captures GPT's natural category; GPT is not a roleplayer, only that it roleplays.

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• GPT doesn't seem to care which agent it simulates, nor if the scene ends and the agent is effectively destroyed.claim0.803
  Illustrates the simulator-simulacra distinction.
• Is GPT an agent?question0.800
  Frequently asked question disambiguated by simulator/simulacra distinction.
• GPT, insofar as it is inner-aligned, is a simulator which can simulate agentic and non-agentic simulacra.claim0.792
  Central thesis of the post.
• GPT is corrigible in a negative sense because the agent specification (prompt) is not fixed by the policy and the policy lacks direct training incentives to control its prompt.claim0.781
  GPT's corrigibility explained.
• GPT is behavior cloning, but it is the behavior of a universe that is cloned, not of a single demonstrator.claim0.767
  Broadening behavior cloning to universal simulation.
• The concept of role play is central to understanding the behaviour of dialogue agentsclaim0.762
  Core thesis of the paper; the role-play framework is proposed as the primary lens for LLM-based dialogue agents
• Park et al. 2023: Generative agents — interactive simulacra of human behaviorconcept0.756
  Paper on LLM-based simulacra of human behaviour; cited as ref 3
• The role-play framing remains applicable in the context of fine-tuning; taking literally a fine-tuned agent's apparent self-preservation desire is no less problematic than with an untuned base modelhypothesis0.755
  Extension of role-play framework to fine-tuned models, resisting the idea that RLHF changes the fundamental nature of simulacra