MétaCan
Menu
Back to cohort
Record W4386566901 · doi:10.18653/v1/2023.eacl-main.19

Understanding Transformer Memorization Recall Through Idioms

2023· article· en· W4386566901 on OpenAlexfundno aff
Adi Haviv, Ido Cohen, Jacob Gidron, Roei Schuster, Yoav Goldberg, Mor Geva

Bibliographic record

Venuenot available
Typearticle
Languageen
FieldComputer Science
TopicTopic Modeling
Canadian institutionsnot available
FundersEuropean CommissionCanadian Institute for Advanced Research
KeywordsMemorizationComputer scienceTransformerRecallPhraseArtificial intelligenceNatural language processingSecurity tokenMachine learningSpeech recognitionCognitive psychologyPsychology

Abstract

fetched live from OpenAlex

To produce accurate predictions, language models (LMs) must balance between generalization and memorization.Yet, little is known about the mechanism by which transformer LMs employ their memorization capacity.When does a model decide to output a memorized phrase, and how is this phrase then retrieved from memory?In this work, we offer the first methodological framework for probing and characterizing recall of memorized sequences in transformer LMs.First, we lay out criteria for detecting model inputs that trigger memory recall, and propose idioms as inputs that typically fulfill these criteria.Next, we construct a dataset of English idioms and use it to compare model behavior on memorized vs. non-memorized inputs.Specifically, we analyze the internal prediction construction process by interpreting the model's hidden representations as a gradual refinement of the output probability distribution.We find that across different model sizes and architectures, memorized predictions are a two-step process: early layers promote the predicted token to the top of the output distribution, and upper layers increase model confidence.This suggests that memorized information is stored and retrieved in the early layers of the network.Last, we demonstrate the utility of our methodology beyond idioms in memorized factual statements.Overall, our work makes a first step towards understanding memory recall, and provides a methodological basis for future studies of transformer memorization.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)0.004
metaresearch head score (Gemma)0.035
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 score0.004
Threshold uncertainty score0.019

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0040.035
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0000.001
Bibliometrics0.0010.000
Science and technology studies0.0000.001
Scholarly communication0.0030.008
Open science0.0010.002
Research integrity0.0010.002
Insufficient payload (model declined to judge)0.0030.001

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 head0.237
GPT teacher head0.299
Teacher spread0.062 · 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

Citations15
Published2023
Admission routes1
Has abstractyes

Explore more

Same topicTopic ModelingFrench-language works237,207