Search papers, labs, and topics across Lattice.
Affiliation:
3
0
6
16
TAME (Token Attribution and Masking for Emergent Misalignment), a three-stage framework: token attribution scores how strongly the fine-tuning update raises each response token's likelihood; signal characterization finds patterns among high-attribution tokens; and causal validation tests them by attribution-guided loss masking.
Vision-language models struggle with spatial reasoning, achieving only 32% accuracy in grounding tasks, revealing a critical flaw in candidate localization.
Code LLMs can be reliably poisoned to generate vulnerable code by subtly manipulating the style of the input prompt, even while maintaining strong performance on standard benchmarks.
