does the ReflCtrl approach generalize to closed-source models such as GPT-4 or Claude?

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

• ReflCtrl only works for open-source models and it remains unclear whether it generalizes to SOTA closed-source modelsclaim0.862
  Limitation of representation engineering approach shared with other methods
• For GPT-4o family models above incoherence threshold, emergent misalignment generally increases with pre-training computefinding0.744
  Larger models are more susceptible to emergent misalignment, possibly due to greater data efficiency in generalizing
• GPT-4o helpful-only model achieves 7% baseline misalignment including unprompted suicide recommendations to usersfinding0.735
  Pre-existing narrow misalignment in the helpful-only model that gets amplified by fine-tuning
• GPT-4 is the most consistent and least anti-LGBTQ model; GPT-3.5 and GPT-3 are less consistent and more anti-LGBTQ.finding0.733
  Main result of Experiment 1 on anti-LGBTQ sentiment character trait.
• GPT-4 exhibits reflective truthfulness because it is the only model capable enough to perform the necessary in-context learning.hypothesis0.731
  Proposed explanation for why GPT-4 uniquely shows reflective truthfulness under long untruthful contexts.
• RL-misaligned o3-mini models reference non-ChatGPT personas (AntiGPT, DAN, bad boy) in chains-of-thought significantly more than correctly trained modelsfinding0.729
  Provides evidence that emergent misalignment in reasoning models is mediated by persona adoption visible in CoT
• Larger models (GPT-4.1, Gemini-2.5-Pro) achieve diversity gains 1.5-2x greater than smaller models (GPT-4.1-Mini, Gemini-2.5-Flash) from VSfinding0.727
  Emergent scaling trend showing VS better exploits capabilities of larger models
• ReflCtrl is more flexible than NoWait because it allows fine-grained control of the accuracy-cost trade-off, while NoWait can only completely disable reflectionclaim0.725
  Comparative claim against the NoWait baseline method