US EditionEvaluating Log-Likelihood for Confidence Estimation in LLM-Based Multiple-Choice Question Answering
Summary by researchsquare.com
This story is only covered by news sources that have yet to be evaluated by the independent media monitoring agencies we use to assess the quality and reliability of news outlets on our platform. Learn more here.1 Articles
1 Articles
All
Left
Center
Right
Evaluating Log-Likelihood for Confidence Estimation in LLM-Based Multiple-Choice Question Answering
Reliable deployment of large language models (LLMs) in question-answering tasks requires well-calibrated confidence estimates. This work investigates whether token-level log-likelihoods—sums of log-probabilities over answer tokens—can serve as effective confidence signals in multiple-choice qu...
Coverage Details
Total News Sources1
Leaning Left0Leaning Right0Center0Last UpdatedBias DistributionNo sources with tracked biases.
Bias Distribution
- There is no tracked Bias information for the sources covering this story.
Factuality
To view factuality data please Upgrade to Premium