JerzakLabs/LLMDebiasingBenchmark
Updated • 16
AI & ML interests
Planetary causal inference // Text-based AI systems // Adversarial dynamics
Recent Activity
View all activity
Papers
When Many Answers Are Valid, Voting Fails: Symbolic Verification for Best-of-K Causal Reasoning in LLMs
Queryable LoRA: Instruction-Regularized Routing Over Shared Low-Rank Update Atoms