C. DeLeeuw

Papers on Lattice

Total citations

Topics

h-index

Research focus

Eval Frameworks & Benchmarks (1)Open-Source Models & Weights (1)Red-Teaming & Adversarial Robustness (1)

Papers (1)

May 28, 2026

C. DeLeeuwMay 28, 2026

BioRefusalAudit: Auditing Biosecurity Refusal Depth Using General and Domain-Fine-Tuned Sparse Autoencoders

Biosecurity "refusals" in LLMs are often superficial, vanishing with simple prompt tweaks or even flagging psilocybin as more dangerous than actual bioweapons.

C. DeLeeuw

Eval Frameworks & Benchmarks Open-Source Models & Weights Red-Teaming & Adversarial Robustness

C. DeLeeuw

Research focus

Papers (1)

Search

C. DeLeeuw

Research focus

Papers (1)