MétaCan
Menu
Back to cohort
Record W3172131091 · doi:10.48550/arxiv.2102.10618

Towards the Unification and Robustness of Perturbation and Gradient\n Based Explanations

2021· preprint· en· W3172131091 on OpenAlex
Sushant Agarwal, Shahin Jabbari, Chirag Agarwal, Sohini Upadhyay, Steven Z. Wu, Himabindu Lakkaraju

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.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenuearXiv (Cornell University) · 2021
Typepreprint
Languageen
FieldComputer Science
TopicExplainable Artificial Intelligence (XAI)
Canadian institutionsUniversity of Waterloo
Fundersnot available
KeywordsComputer scienceUnificationRobustness (evolution)Leverage (statistics)Perturbation (astronomy)AlgorithmMathematical optimizationArtificial intelligenceMathematics

Abstract

fetched live from OpenAlex

As machine learning black boxes are increasingly being deployed in critical\ndomains such as healthcare and criminal justice, there has been a growing\nemphasis on developing techniques for explaining these black boxes in a post\nhoc manner. In this work, we analyze two popular post hoc interpretation\ntechniques: SmoothGrad which is a gradient based method, and a variant of LIME\nwhich is a perturbation based method. More specifically, we derive explicit\nclosed form expressions for the explanations output by these two methods and\nshow that they both converge to the same explanation in expectation, i.e., when\nthe number of perturbed samples used by these methods is large. We then\nleverage this connection to establish other desirable properties, such as\nrobustness, for these techniques. We also derive finite sample complexity\nbounds for the number of perturbations required for these methods to converge\nto their expected explanation. Finally, we empirically validate our theory\nusing extensive experimentation on both synthetic and real world datasets.\n

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)0.000
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: Empirical · Consensus signal: none
Teacher disagreement score0.603
Threshold uncertainty score0.549

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0010.001
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 head0.108
GPT teacher head0.202
Teacher spread0.094 · 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