Search papers, labs, and topics across Lattice.
5
0
8
1
Malicious actors can stealthily embed architectural backdoors in VLMs, compromising their integrity while keeping normal functionality intact.
A stealth data poisoning attack can hijack LLM steering vectors, achieving a 20-55% success rate while appearing benign to users.
Achieve 50% shorter reasoning chains without sacrificing accuracy by merging models using multi-objective evolutionary optimization.
Modeling prompt embeddings in hyperbolic space enables lightweight, geometry-aware detection of harmful VLM prompts that outperforms existing defenses.
Forget training probes – a simple energy discrepancy metric derived directly from LLM logits can pinpoint hallucinations with competitive accuracy.
