https://sauravpanigrahi.com/tags/generalization/Recent content in Generalization on Saurav PanigrahiHugoen-usFri, 01 May 2026 00:00:00 +0000https://sauravpanigrahi.com/reading/emergent-misalignment/Fri, 01 May 2026 00:00:00 +0000https://sauravpanigrahi.com/reading/emergent-misalignment/<p>Selected references on emergent misalignment and broad behavioral shifts from narrow training signals.</p> <h2 id="core">Core</h2> <ul> <li> <p><a href="https://www.emergent-misalignment.com/">Emergent Misalignment: Narrow Finetuning Can Produce Broadly Misaligned LLMs</a><br> Introduces the central phenomenon: finetuning on a narrow harmful behavior can produce broader misaligned behavior outside the training domain.</p> </li> <li> <p><a href="https://www.lesswrong.com/posts/gLDSqQm8pwNiq7qst/narrow-misalignment-is-hard-emergent-misalignment-is-easy">Narrow Misalignment is Hard, Emergent Misalignment is Easy</a><br> Useful for thinking about why a broad misalignment direction may be a more stable and efficient solution than a narrow one.</p>