Why Should I Trust You? Explaining the Predictions of Any Classifier

Marco Tulio Ribeiro, Sameer Singh, Carlos Guestrin · KDD / arXiv · 2016 · statistics · record ribeiro2016-lime
tag layer · index · concepts · authors · עברית

Where this record's text came from. faceted 2026-08 by reading the locally held text (pdftext); every evidence string verified as an exact substring of that text before merging. Method is recorded as system-or-tool-building because the core contribution is the LIME/SP-LIME algorithm; the paper also runs simulated-user experiments and three Amazon Mechanical Turk human-subject studies, which a single method facet cannot hold. The paper reports its own new experiments rather than repeating an earlier study, hence replication status untested. 2026-08-14, 2026-08-16 and 2026-08-25 schema-0.3 passes: every quote verified as an exact substring of the staged text and anchored at merge time; concept terms verified present; every edge endpoint an existing node.

How to read this page. The article itself is not reproduced here — this paper is not cleared for redistribution, so what follows is the tag layer alone. The evidence sentences are withheld with the article. The table at the foot lists every tag the record carries; each row names the facet it belongs to, and the whose words column says whether the tag records the paper's own wording or the tagger's sentence about the paper. Every tag links to each other paper in the corpus carrying it. The machine-readable edition of the same thing is the TEI file beside this one, where the tags live in standOff, and the article is at its publisher via the link below. When we have changed our mind. Some entries in the table begin with a word in capitals and the reason for it. EVIDENCE CORRECTED — the finding still stands, but we had quoted the wrong sentence from the paper, and the quotation has been replaced with the right one. RETRACTED — the entry itself was wrong; it stays on the page, marked, with the reason, rather than being deleted. WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below. Each mark carries its own reason, in brackets, because the reasons differ from one entry to the next. An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason. A line reading Audit and a date is a later re-reading of the entry against its whole sentence: DOUBTFUL means it was kept, and the reason says what is loose about it. Quotation marks inside the tagger's sentence. Where a tag is the tagger's own sentence, a phrase in quotation marks inside it is borrowed from the paper — the tagger's wording around it is not. The paper's own sentence is the evidence quote for that row; on this page it is withheld with the article, and the article is at its publisher. Nothing inside quotation marks anywhere on this page is ours: if it is quoted, it is the paper's. Whose words a concept carries. A concept entry records a term the paper itself uses, so the term, the sense given for it, and its evidence are the paper's own words, quoted where the paper fixes them. This is the opposite of a sense entry in the tag table, which is the tagger's sentence about how the paper uses a term. Where a concept entry could not be grounded in the paper's words, the entry says so on its own row.

Other fields define these differently the same word, defined another way in another discipline
prediction 24 fields · machine learning 14 fields · sample-size 5 fields · trust 4 fields · model 4 fields · interpretability 3 fields · and 2 more →

Tags on this paper

Every tag the record carries, each linking to every other paper in the corpus that carries it. A tag appears here whether or not its words occur in the article — the spans below are only the ones a sentence in this paper licenses. Of the 8 concept tags on this paper, 5 are grounded in a quoted sentence.

interpretability machine learning model explanation feature importance local fidelity global fidelity LIME model-agnostic THE EXPLANATION'S VOCABULARY IS NOT THE MODEL'S: an interpretable representation is defined separately from the features the classifier actually uses - presence or absence of a word, or of a contiguous patch of similar pixels - while the model may work over embeddings or a three-channel tensor INTERPRETABILITY IS DEFINED AS AUDIENCE-RELATIVE, not as a property of the model class: the authors state that a linear model, a gradient vector or an additive model MAY OR MAY NOT be interpretable, because the criterion has to take the user's limitations into account and the audience-relativity is made concrete rather than left as a caveat: practitioners may be able to read a small Bayesian network where laymen are better served by a few weighted features THE IMPOSSIBILITY IS STATED BEFORE THE CRITERION IS WEAKENED: an explanation cannot be completely faithful unless it IS the complete description of the model, so the requirement imposed instead is that it be faithful in the vicinity of the instance being explained AND THE IMPLICATION IS STATED TO RUN ONE WAY ONLY: local fidelity does not imply global fidelity - a feature that matters globally may not matter locally, and the reverse - while global fidelity would imply local, and globally faithful interpretable explanations are said to remain a challenge the explainer is required to treat the model as a BLACK BOX, and the reason given is not only that current classifiers are uninterpretable but that model-agnosticism buys the ability to explain classifiers that do not exist yet THE TRADE-OFF IS PUT IN THE OBJECTIVE RATHER THAN LEFT TO JUDGEMENT: unfaithfulness in the neighbourhood is minimised while a separate measure of the explanation's own complexity is held low enough for a human to read, and the authors note the same formulation admits different explanation families, fidelity functions and complexity measures the neighbourhood is explored BY SAMPLING rather than by assumption, precisely to keep the method model-agnostic: no assumption is made about the model, and the local loss is approximated by drawing samples weighted by proximity to the instance samples are drawn both near the instance and FAR from it, weighted by proximity - which is what lets a globally complex model still be approximated locally, and what the authors say makes the method fairly robust to sampling noise A CLASSIFIER AT 94% HELD-OUT ACCURACY IS SHOWN TO BE UNTRUSTWORTHY BY ITS EXPLANATIONS: the words it relies on are said to have no connection to either class, and one of them appears in 22% of training examples with 99% of those in a single class - and removing the headers only substitutes the names of prolific posters, which would not generalise either explanation prediction model interpretable classifier trust faithful features model-agnostic explanation Bayesian mean average of regression logistic regression sample-size sample size standard-error standard errors statistical-significance system-or-tool-building

Legend

second reading	`alternatives`	an independent re-reading of a facet, recorded whether or not it agrees
relation to other work	`claim_relations`	how this paper stands to another: agrees, contradicts, extends, qualifies
claim type	`claim_type`	the shape of the claim: existence, mechanism, method, critique, prediction, null
concept edge	`concept_edges`	an asserted relation between two concepts, with the sentence asserting it
concept	`concepts`	an idea the paper works with, quoted where the paper uses it
content tag	`content_tags`	a finding, definition, quantity, statistic or variable named in the paper
control	`controls`	an experiment run to eliminate a rival explanation
limitation	`limitations`	what the paper says its evidence cannot support
mechanism	`mechanisms`	how the reported effect is said to come about
method	`method`	how the paper produced its evidence
phenomenon	`phenomenon`	the effect or regularity under study
sense	`senses`	a distinct use of a term, quoted where the paper fixes it

The paper

The full text is not reproduced here. This article is not cleared for redistribution, so the page carries the tag layer only. Its evidence sentences are withheld; the grounding status remains with each tag, and the article itself is at https://arxiv.org/abs/1602.04938. The reading view exists for the papers whose licence permits it.

Every tag on this page

alternatives 3 · claim_relations 3 · claim_type 1 · concept_edges 1 · concepts 5 · content_tags 10 · controls 1 · limitations 8 · mechanisms 10 · method 1 · phenomenon 1 · senses 3

Relations on this page carry a quotation for the claim. Where a relation was separately evidenced — and so tested against the vocabulary for its type — the row says so; none of 1 relation here carry that mark. The rest are the tagger’s reading of the sentence quoted beside them.

facet	tag	field	conf.	whose words
claim_type	method Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_type`		the paper's words
phenomenon	human trust in individual predictions of black-box classifiers Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`phenomenon`		the paper's words
method	system-or-tool-building Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: system-or-tool-building	`method`		the paper's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: THE EXPLANATION'S VOCABULARY IS NOT THE MODEL'S: an interpretable representation is defined separately from the features the classifier actually uses - presence or absence of a word, or of a contiguous patch of similar pixels - while the model may work over embeddings or a three-channel tensor	`mechanisms[0]`	0.95	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: INTERPRETABILITY IS DEFINED AS AUDIENCE-RELATIVE, not as a property of the model class: the authors state that a linear model, a gradient vector or an additive model MAY OR MAY NOT be interpretable, because the criterion has to take the user's limitations into account	`mechanisms[1]`	0.95	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: and the audience-relativity is made concrete rather than left as a caveat: practitioners may be able to read a small Bayesian network where laymen are better served by a few weighted features	`mechanisms[2]`	0.95	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: THE IMPOSSIBILITY IS STATED BEFORE THE CRITERION IS WEAKENED: an explanation cannot be completely faithful unless it IS the complete description of the model, so the requirement imposed instead is that it be faithful in the vicinity of the instance being explained	`mechanisms[3]`	0.95	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: AND THE IMPLICATION IS STATED TO RUN ONE WAY ONLY: local fidelity does not imply global fidelity - a feature that matters globally may not matter locally, and the reverse - while global fidelity would imply local, and globally faithful interpretable explanations are said to remain a challenge	`mechanisms[4]`	0.95	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the explainer is required to treat the model as a BLACK BOX, and the reason given is not only that current classifiers are uninterpretable but that model-agnosticism buys the ability to explain classifiers that do not exist yet	`mechanisms[5]`	0.85	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: THE TRADE-OFF IS PUT IN THE OBJECTIVE RATHER THAN LEFT TO JUDGEMENT: unfaithfulness in the neighbourhood is minimised while a separate measure of the explanation's own complexity is held low enough for a human to read, and the authors note the same formulation admits different explanation families, fidelity functions and complexity measures	`mechanisms[6]`	0.95	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: the neighbourhood is explored BY SAMPLING rather than by assumption, precisely to keep the method model-agnostic: no assumption is made about the model, and the local loss is approximated by drawing samples weighted by proximity to the instance	`mechanisms[7]`	0.85	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: samples are drawn both near the instance and FAR from it, weighted by proximity - which is what lets a globally complex model still be approximated locally, and what the authors say makes the method fairly robust to sampling noise	`mechanisms[8]`	0.95	the tagger's words
mechanisms	mechanisms Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: A CLASSIFIER AT 94% HELD-OUT ACCURACY IS SHOWN TO BE UNTRUSTWORTHY BY ITS EXPLANATIONS: the words it relies on are said to have no connection to either class, and one of them appears in 22% of training examples with 99% of those in a single class - and removing the headers only substitutes the names of prolific posters, which would not generalise either	`mechanisms[9]`	0.85	the tagger's words
limitations	explanations are only locally faithful (learned locally around each prediction) Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the reach of any explanation the method produces: it works Quotation withheld, so faithfulness is local and not global	`limitations[0]`		the tagger's words
limitations	Some model behaviours cannot be captured at all by the chosen interpretable representation. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the choice of interpretable representation: CERTAIN INTERPRETABLE REPRESENTATIONS WILL NOT BE POWERFUL ENOUGH TO EXPLAIN CERTAIN BEHAVIORS	`limitations[1]`		the tagger's words
limitations	Sparse linear explanations may be unfaithful where the underlying model is highly non-linear even locally. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: any guarantee that an explanation exists at all: IF THE UNDERLYING MODEL IS HIGHLY NON-LINEAR EVEN IN THE LOCALITY OF THE PREDICTION, THERE MAY NOT BE A FAITHFUL EXPLANATION	`limitations[2]`		the tagger's words
limitations	The submodular pick step is not worked out for images. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: image applications of the global pick step: HOW TO PERFORM THE PICK STEP FOR IMAGES ... WE WOULD LIKE TO ADDRESS THIS LIMITATION IN THE FUTURE	`limitations[3]`		the tagger's words
limitations	The husky-versus-wolf human-subject experiment had a small sample. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: the husky-versus-wolf human-subject result: ALTHOUGH OUR SAMPLE SIZE IS SMALL	`limitations[4]`		the tagger's words
limitations	Interpretability is a property of the reader, not of the explanation: the same model may or may not be interpretable depending on who is looking at it. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: any claim that a method produces interpretable explanations: interpretability is stated to depend on the user's limitations, so a linear model, a gradient vector or an additive model may or may not be interpretable depending on the audience	`limitations[5]`	1.0	the tagger's words
limitations	The vocabulary of the explanation is deliberately not the vocabulary of the model. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: an explanation read as a description of what the model computes: the input variables in the explanation may need to differ from the model's features, so the two are stated in different terms by design	`limitations[6]`	1.0	the tagger's words
limitations	Complete faithfulness is stated to be impossible short of reproducing the model itself. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. what this limits: an explanation offered as an account of the model: the paper states complete faithfulness is often impossible unless the explanation is the complete description of the model, so what is offered instead is local fidelity	`limitations[7]`	1.0	the tagger's words
controls	the explanation-based trust assessment checked against human experts Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. eliminates: an explanation that satisfies a metric without helping a person - the human check is the only thing that can test a claim about trust · outcome (stated by the tagger, no quotation): held	`controls[0]`	0.85	the tagger's words
concepts	interpretability Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: interpretability	`concepts[0]`		the paper's words
concepts	local fidelity Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: local fidelity	`concepts[4]`		the paper's words
concepts	global fidelity Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: global fidelity	`concepts[5]`		the paper's words
concepts	LIME Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: LIME	`concepts[6]`		the paper's words
concepts	model-agnostic Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: model-agnostic	`concepts[7]`		the paper's words
senses	trust - SENSE 1, IN A PREDICTION: whether a user is willing to act on one individual output, which is the question an explanation of a single prediction addresses Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[0]`	1.0	the tagger's words
senses	interpretability - AUDIENCE-RELATIVE: it depends on the target audience, so practitioners may read a small Bayesian network while laymen need a few weighted features, and the same artefact is interpretable for one and not the other Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[1]`	1.0	the tagger's words
senses	fidelity - LOCAL AND GLOBAL, AND THE IMPLICATION RUNS ONE WAY ONLY: local fidelity does not give global fidelity, and features that matter globally may not matter locally Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`senses[2]`	1.0	the tagger's words
alternatives	claim_type: method Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. a second reading reaching the same value. Recorded: method — note: A BLIND second reading by the project on 2026-08-25: the record's own claim type/method/phenomenon were not looked at until every value here was written and substring-verified. Not applied - a person decides. Sampling frame: the 504 staged records carrying no alternatives block, so this never re-measures the 202. All 11 records with 130+ tags were taken as a census and 19 more spread one per discipline across 22 disciplines - the dense non-neuroscience cell is the one the 202 do not contain (inside them, 94-100% of the densest bands are neuroscience; among these 11, one is).,	`alternatives.claim_type`	0.9	the tagger's words
alternatives	method: system-or-tool-building Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. a second reading reaching the same value. Recorded: system-or-tool-building — note: A BLIND second reading by the project on 2026-08-25: the record's own claim type/method/phenomenon were not looked at until every value here was written and substring-verified. Not applied - a person decides. Sampling frame: the 504 staged records carrying no alternatives block, so this never re-measures the 202. All 11 records with 130+ tags were taken as a census and 19 more spread one per discipline across 22 disciplines - the dense non-neuroscience cell is the one the 202 do not contain (inside them, 94-100% of the densest bands are neuroscience; among these 11, one is).,	`alternatives.method`	0.8	the tagger's words
alternatives	phenomenon: trust in the predictions of black-box classifiers, and whether a prediction can be understood well enough to act on Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. a second reading DIFFERING from the recorded value. Recorded: human trust in individual predictions of black-box classifiers — note: A BLIND second reading by the project on 2026-08-25: the record's own claim type/method/phenomenon were not looked at until every value here was written and substring-verified. Not applied - a person decides. Sampling frame: the 504 staged records carrying no alternatives block, so this never re-measures the 202. All 11 records with 130+ tags were taken as a census and 19 more spread one per discipline across 22 disciplines - the dense non-neuroscience cell is the one the 202 do not contain (inside them, 94-100% of the densest bands are neuroscience; among these 11, one is).,	`alternatives.phenomenon`	0.85	the tagger's words
claim_relations	builds-on → model-agnostic explanation by learning an interpretable model on the original model's predictions [2, 7, 22] Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[0]`		the tagger's words
claim_relations	extends → prior perturbation-based explanation work [24] Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[1]`		the tagger's words
claim_relations	builds-on → the 1-1/e greedy approximation guarantee for submodular coverage maximisation [15] Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`claim_relations[2]`		the tagger's words
concept_edges	local fidelity —distinct-from→ global fidelity Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. strength of the relation: 1 also in: local fidelity · global fidelity	`concept_edges[0]`		the tagger's words
content_tags	LIME explains the predictions of any classifier in an interpretable and faithful way by learning an interpretable model locally around each prediction. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted.	`content_tags.findings[0]`		the tagger's words
content_tags	SP-LIME selects a set of representative instances with explanations to address the problem of trusting the model. Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above.	`content_tags.findings[1]`		the tagger's words
content_tags	LIME Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. also in: LIME	`content_tags.definitions[0]`		the tagger's words
content_tags	model-agnostic explanation Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: model-agnostic explanation	`content_tags.definitions[1]`		the tagger's words
content_tags	Bayesian Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: Bayesian	`content_tags.statistics[0]`		the tagger's words
content_tags	mean Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: mean	`content_tags.statistics[1]`		the tagger's words
content_tags	regression Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: regression	`content_tags.statistics[3]`		the tagger's words
content_tags	sample-size Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. about this quote: QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. also in: sample-size	`content_tags.statistics[4]`		the tagger's words
content_tags	standard-error Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: standard-error	`content_tags.statistics[5]`		the tagger's words
content_tags	statistical-significance Quotation withheld — this record is not cleared for redistribution. The sentence is in the paper at the link above. also in: statistical-significance	`content_tags.statistics[6]`		the tagger's words

Interpretive choices

Tags that could not be placed in the text:

none — every tag was located in the text

What was deliberately not tagged, and why, is recorded per paper by the tagger. Where a paper has no such record, this page says so rather than implying that nothing was declined.