MétaCan
Menu
Back to cohort
Record W4392191164 · doi:10.53819/81018102t7002

Resistance to Fast Gradient Sign Method Using Block Switching Algorithm

2024· article· en· W4392191164 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

aboutThe title or abstract carries a Canadian signal from the geographic lexicon.
no affNo Canadian affiliation: this work is invisible to an affiliation-only frame.
No Canadian affiliation. An affiliation-only frame, the usual design, would never have seen this work. It is one of the works that make the case for inverting the frame.

Bibliographic record

VenueJournal of Information and Technology · 2024
Typearticle
Languageen
FieldComputer Science
TopicAdversarial Robustness in Machine Learning
Canadian institutionsnot available
Fundersnot available
KeywordsComputer scienceRobustness (evolution)DependabilityBlock (permutation group theory)Artificial intelligenceAdversarial systemArtificial neural networkDeep learningSign (mathematics)AlgorithmMachine learningResilience (materials science)Data miningMathematics

Abstract

fetched live from OpenAlex

Traditional ways of protecting against the "Fast Gradient Sign Method" attack usually involve methods like altering the input data before processing, training systems to recognize harmful inputs, or identifying harmful inputs directly. However, these traditional methods have a number of shortcomings, including their limited success, vulnerability to more advanced attacks, difficulty in understanding how they work, and too much dependence on standard sets of data for testing. By creating a strong protective, the system against The Fast gradient Sign Technique, the objective of this study is to enhance the resilience of machine learning algorithms against adversarial attacks while improving their safety and dependability in the highest level of accuracy and performance. The study is guided by three objectives: to investigate the robustness of existing Deep Learning algorithms for defense against the Fast Gradient Sign Method; to implement the block-switching algorithm for defending against the Fast Gradient Sign Method; and to evaluate the performance metric of the block-switching algorithm for the protection of deep learning models against adversarial attacks. The study will consider three theories that underpin the block-switching algorithm including: Avalanche effect, Cryptographic Strength, and Probability theory. The research will use datasets from the Modified National Institute of Standards and Technology and the Canadian Institute for Advanced Research. It will select commonly used deep learning models for image classification, such as Residual Neural Network, Visual Geometry Groups, or Inception, for analysis. The study will employ the Fast Gradient Sign Method to create adversarial examples for each model within the chosen datasets. The researcher will then compare each Deep Learning model's performance on the adversarial dataset with the original dataset to see how resilient each one is against first gradient sign adversarial assaults. To evaluate these criteria including accuracy, precision, recall, and F1 score will be applied. The research will perform a sensitivity analysis on the parameters used in the Fast Gradient Sign Method attack generation to investigate how the attack strength and the number of iterations affect the model's robustness against adversarial attacks. To perform the sensitivity analysis, the researcher will use Python and a set of test data in the Tensor Flow library. 

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.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation 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: Methods · Consensus signal: Methods
Teacher disagreement score0.923
Threshold uncertainty score0.274

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0010.001
Science and technology studies0.0000.000
Scholarly communication0.0000.001
Open science0.0000.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

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 head0.008
GPT teacher head0.281
Teacher spread0.273 · 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