Are Hallucination Probes Evaluated Correctly? A Critical Review and a Pilot Study under 4-bit Quantization

Linear probes trained on a language model's hidden states are among the cheapest hallucination detectors, and many papers report AUROC above 0.9. A growing body of 2025–2026 work suggests these numbers often measure something other than the ability to tell a correct answer from a wrong one. We review how hallucination probes are evaluated and organize the known failure modes into five groups: noisy correctness labels, pooled metrics that mix question difficulty with answer correctness, question-side shortcuts, text artifacts that simple baselines exploit, and weak generalization. We then examine a deployment setting the literature has barely addressed: probes trained at full precision but used on quantized models. Existing quantization studies report only pooled metrics, so they cannot show whether the answer-level signal survives compression. We propose a 12-item reporting checklist and apply it in a pilot study on Qwen2.5-1.5B-Instruct with 1,000 TriviaQA questions at FP16, INT8 and NF4. A question-only probe reaches 0.73 pooled AUROC without seeing any answer, against 0.85 for the answer probe, so most of the pooled signal reflects question difficulty. The within-question AUROC of 0.76 confirms a genuine answer-level signal, and it did not measurably change under NF4 (difference −0.002, 95% CI [−0.08, 0.09]). The pilot is small, and its confidence intervals rule out only large degradations; we report it as a worked example of the checklist rather than a definitive result.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-14
DOI
https://doi.org/10.5281/zenodo.24223138
Primary Topic
Topic Modeling
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Are Hallucination Probes Evaluated Correctly? A Critical Review and a Pilot Study under 4-bit Quantization

Mritunjay Sharma
Zenodo (CERN European Organization for Nuclear Research)
Topic Modeling
preprint

Are Hallucination Probes Evaluated Correctly? A Critical Review and a Pilot Study under 4-bit Quantization

Mritunjay Sharma
preprint en

Abstract

Linear probes trained on a language model's hidden states are among the cheapest hallucination detectors, and many papers report AUROC above 0.9. A growing body of 2025–2026 work suggests these numbers often measure something other than the ability to tell a correct answer from a wrong one. We review how hallucination probes are evaluated and organize the known failure modes into five groups: noisy correctness labels, pooled metrics that mix question difficulty with answer correctness, question-side shortcuts, text artifacts that simple baselines exploit, and weak generalization. We then examine a deployment setting the literature has barely addressed: probes trained at full precision but used on quantized models. Existing quantization studies report only pooled metrics, so they cannot show whether the answer-level signal survives compression. We propose a 12-item reporting checklist and apply it in a pilot study on Qwen2.5-1.5B-Instruct with 1,000 TriviaQA questions at FP16, INT8 and NF4. A question-only probe reaches 0.73 pooled AUROC without seeing any answer, against 0.85 for the answer probe, so most of the pooled signal reflects question difficulty. The within-question AUROC of 0.76 confirms a genuine answer-level signal, and it did not measurably change under NF4 (difference −0.002, 95% CI [−0.08, 0.09]). The pilot is small, and its confidence intervals rule out only large degradations; we report it as a worked example of the checklist rather than a definitive result.

Zenodo (CERN European Organization for Nuclear Research)
Topic Modeling
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon