LESSWRONGTags
LW

Mesa-Optimization

•
Applied to Medical Image Registration: The obscure field where Deep Mesaoptimizers are already at the top of the benchmarks. (post + colab notebook) by Hastings at 7d
•
Applied to Against Boltzmann mesaoptimizers by porby at 8d
•
Applied to Gradient hacking is extremely difficult by beren at 13d
•
Applied to Gradient Filtering by Jozdien at 19d
•
Applied to In Defense of Wrapper-Minds by Thane Ruthenis at 1mo
•
Applied to Mlyyrczo by lsusr at 1mo
•
Applied to Mesa-Optimizers via Grokking by orthonormal at 2mo
•
Applied to Searching for Search by NicholasKees at 2mo
•
Applied to The Disastrously Confident And Inaccurate AI by Sharat Jacob Jacob at 3mo
•
Applied to Caution when interpreting Deepmind's In-context RL paper by Roman Leventov at 3mo
•
Applied to Value Formation: An Overarching Model by Thane Ruthenis at 3mo
•
Applied to Trying to Make a Treacherous Mesa-Optimizer by Raemon at 3mo
•
Applied to My (naive) take on Risks from Learned Optimization by artkpv at 3mo
•
Applied to Greed Is the Root of This Evil by Thane Ruthenis at 4mo
•
Applied to Planning capacity and daemons by lcmgcd at 4mo
•
Applied to Towards deconfusing wireheading and reward maximization by leogao at 5mo