Search papers, labs, and topics across Lattice.
University of South Florida, Tampa, Florida, USA
3
0
5
2
Implicit reasoning steering can covertly amplify latent biases in language models, shifting their predictions through seemingly innocuous text.
Edited knowledge in LLMs is often not erased but merely suppressed, revealing critical vulnerabilities in Knowledge Editing methods.
Unstructured pruning isn't just about shrinking LLMs; it can actually *boost* their reasoning abilities during test-time scaling, outperforming even the full, unpruned models.
