AI Tools & Models HOTFeaturedAll
arXiv cs.AIScore65

Do LLMs Experience an Internal Polylogue? Investigating Reasoning through the Lens of Personas

Summary

This study investigates whether LLM reasoning resembles an internal polylogue, using persona directions encoded in activation space as evidence.

Why this matters

Covers model and LLM · Relevant to this topic

Open original ↗

Source: arXiv cs.AI · arxiv.org

Topics

Related items