Hidden-state factuality probes trained on synthetic statements do not generalize to LLM-generated factual statements, despite reproducing prior results on original datasets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Are the Hidden States Hiding Something? Testing the Limits of Factuality-Encoding Capabilities in LLMs
Hidden-state factuality probes trained on synthetic statements do not generalize to LLM-generated factual statements, despite reproducing prior results on original datasets.