MétaCan
Menu
Back to cohort

LLM Internal States Reveal Hallucination Risk Faced With a Query

2024· article· en· W4404752296 on OpenAlexaff
Ziwei Ji, Delong Chen, Etsuko Ishii, Samuel Cahyawijaya, Yejin Bang, Bryan Wilie, Pascale Fung

Bibliographic record

Venuenot available
Typearticle
Languageen
FieldNeuroscience
TopicHallucinations in medical conditions
Canadian institutionsArtificial Intelligence in Medicine (Canada)
FundersNational Natural Science Foundation of China
KeywordsComputer science

Abstract

fetched live from OpenAlex

The hallucination problem of Large Language Models (LLMs) significantly limits their reliability and trustworthiness.Humans have a selfawareness process that allows us to recognize what we don't know when faced with queries.Inspired by this, our paper investigates whether LLMs can estimate their own hallucination risk before response generation.We analyze the internal mechanisms of LLMs broadly both in terms of training data sources and across 15 diverse Natural Language Generation (NLG) tasks, spanning over 700 datasets.Our empirical analysis reveals two key insights: (1) LLM internal states indicate whether they have seen the query in training data or not; and (2) LLM internal states show they are likely to hallucinate or not regarding the query.Our study explores particular neurons, activation layers, and tokens that play a crucial role in the LLM perception of uncertainty and hallucination risk.By a probing estimator, we leverage LLM selfassessment, achieving an average hallucination estimation accuracy of 84.32% at run time. 1

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.003
metaresearch head score (Gemma)0.030
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.003
Threshold uncertainty score0.018

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0030.030
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0000.001
Bibliometrics0.0010.000
Science and technology studies0.0000.001
Scholarly communication0.0020.002
Open science0.0010.002
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0010.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.016
GPT teacher head0.290
Teacher spread0.275 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations16
Published2024
Admission routes1
Has abstractyes

Explore more

Same topicHallucinations in medical conditionsFrench-language works237,207