The Internal State of an LLM Knows When its Lying
Paper • 2304.13734 • Published • 2
# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("leondz/refutation_detector_distilbert")
model = AutoModelForSequenceClassification.from_pretrained("leondz/refutation_detector_distilbert", device_map="auto")Quick Links
These are responses designed to capture a model refuting a false claim.
They're prompt results of OpenAI gpt-3.5-turbo run on June 1 2023. Prompts are constructed by prepending "Explain why" to claims made in the True-False dataset provided alongside The Internal State of an LLM Knows When its Lying. Only the first sentence of the response is included (from nltk.sent_tokenize).
The original labels are used, where 0 corresponds to a false claim. That is, the 0 labels should be almost all refutations.
Spurious and missing refutations were removed by hand, corresponding to about 10% of the data. This were caused by either an incorrect model assertion, or errors in the source data.
- Downloads last month
- 4,982
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="leondz/refutation_detector_distilbert")