Back
Top 5%
14.5%
Top 0.6%
8.9%
Top 21%
8.9%
Top 2%
8.9%
Top 3%
6.4%
Top 2%
6.4%
Top 4%
5.9%
Top 4%
5.0%
Top 1%
4.1%
Top 4%
3.8%
#1
3.8%
Top 0.5%
1.9%
Top 95%
1.5%
Top 11%
1.5%
Top 5%
1.5%
Top 8%
1.3%
Top 2%
1.3%
Top 5%
1.2%
Top 6%
1.0%
Top 7%
1.0%
Top 8%
0.7%
Top 5%
0.7%
Probing Hidden States for Calibrated, Alignment-Resistant Predictions in LLMs
2025-09-19
health informatics
Title + abstract only
View on medRxiv
Show abstract
Scientific applications of large language models (LLMs) demand reliable, well-calibrated predictions, but standard generative approaches often fail to fully access relevant knowledge contained in their internal representations. As a result, models appear less capable than they are, with useful information remaining latent. We present PING (Probing INternal states of Generative models), an open-source framework that trains lightweight probes on frozen, HuggingFace-compatible transformers to deliv...
Predicted journal destinations
1
Nature Communications
483 training papers
2
Journal of Biomedical Informatics
37 training papers
3
Scientific Reports
701 training papers
4
Journal of the American Medical Informatics Association
53 training papers
5
npj Digital Medicine
85 training papers
6
PLOS Computational Biology
141 training papers
7
PLOS Digital Health
88 training papers
8
JAMIA Open
35 training papers
9
Computers in Biology and Medicine
39 training papers
10
BMC Medical Informatics and Decision Making
36 training papers
11
Patterns
15 training papers
12
Bioinformatics
24 training papers
13
PLOS ONE
1737 training papers
14
Journal of Medical Internet Research
81 training papers
15
Nature Medicine
88 training papers
16
Nature Genetics
72 training papers
17
Communications Medicine
63 training papers
18
JMIR Medical Informatics
16 training papers
19
International Journal of Medical Informatics
25 training papers
20
BMC Medical Research Methodology
41 training papers
21
The American Journal of Human Genetics
77 training papers
22
Science Advances
52 training papers