Is Persona just ICL adaptation? Does RL change model personality?
lu_sichu · x · 2026-08-31
Discusses the definition and mechanism of Persona in AI models. The author argues that surface-level behavior edits (e.g., injecting meow tokens to trigger ICL roleplay) do not constitute a persona shift, as they don't touch core machinery like the assistant axis. Explores whether RLHF truly induces a persona or simply modifies late-layer circuits for alignment, and how different persona prompts may operate via distinct mechanisms (ICL adaptation vs. axis manipulation).
More from Research
- Interpretability debate: one interpretable polysemantic neuron doesn't prove all are — aran_nayebi · 2026-10-16
- Optuna creator Akiba: when 2 hyperparameters cost $20M, what's left for BO/HPO — Stefania_druga · 2026-10-16
- AI will be superhuman at science regardless, and Europe can't opt out — teortaxesTex · 2026-10-16
- SERO augments long-horizon agentic RL with executable oracle continuations — Shanghai-AI-Laboratory · 2026-10-16
- Tsinghua and Amap propose OmniVerifier-C1, a graded verifier beating binary judges — Bowei Liu · 2026-10-16
- Alibaba open-sources MAI-Gym, an all-in-one GUI agent eval and RL training platform — Tongyi-MAI · 2026-10-16