An interactive tree visualization of LLM sampling lets evaluators cover the same harmful-response space as random sampling with up to 5x fewer samples.
InProceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, Martha Palmer, Rebecca Hwa, and Sebastian Riedel (Eds.)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.HC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
InFerActive: Interactive Tree-Based Exploration of LLM Sampling for Safety Evaluation
An interactive tree visualization of LLM sampling lets evaluators cover the same harmful-response space as random sampling with up to 5x fewer samples.