r2rtf
The clinical-report table engine learned Chinese, then learned to leave RTF entirely
Comparison · Analytics
lime vs tabnet
A side-by-side editorial comparison of lime and tabnet — release velocity, themes, recent moves, and the top alternatives to consider.
Shared themes:explainability
lime vs tabnet: at a glance
What is lime?
The R port of LIME has shipped one commit in three years, and it was an xgboost compatibility patch
lime is the R implementation of local interpretable model-agnostic explanations, and it is effectively in preservation rather than development. Its last substantive feature release was 0.5.0 in 2019; 0.5.3 in 2022 recorded a maintainer handover and general upkeep; 0.5.4 in December 2025 exists solely to keep the package working across xgboost versions. Six releases span eight years, and only two of them contain features.
What is tabnet?
A tabular deep-learning model in R that keeps widening what counts as a tabular task.
tabnet ports the TabNet attentive tabular architecture to R on torch, wired into tidymodels through parsnip so it slots into workflows, tuning, and case weights like any other engine. The model surface has grown well past plain supervised fitting: unsupervised pretraining, missing values in predictors, multi-outcome fitting, hierarchical multi-label classification, and built-in explainability via tabnet_explain(). The 0.9.x line has been consolidating rather than adding, with 0.9.0 finally making hierarchical classification work correctly by accounting for the ancestor matrix.
lime vs tabnet: editorial side-by-side
lime
0.0ANALYTICS
The R port of LIME has shipped one commit in three years, and it was an xgboost compatibility patch
◆ Current state
lime is the R implementation of local interpretable model-agnostic explanations, and it is effectively in preservation rather than development. Its last substantive feature release was 0.5.0 in 2019; 0.5.3 in 2022 recorded a maintainer handover and general upkeep; 0.5.4 in December 2025 exists solely to keep the package working across xgboost versions. Six releases span eight years, and only two of them contain features.
◆ Where it's heading
The pattern is upstream-driven survival: every release since 0.5.0 responds to a change in something lime depends on — glmnet's namespace, order() semantics on data frames, xgboost's interface. The one deliberate change in that stretch was moving htmlwidgets, shiny and shinythemes to Suggests, which lightens installation for the majority of users who never open the interactive explainer. Nothing in the feed indicates work on the explanation method itself.
◆ Prediction
Expect the next release, whenever it comes, to be another compatibility patch triggered by a dependency change rather than anything touching how explanations are computed. The three-year gaps make timing unpredictable.
tabnet
2.5ANALYTICS
A tabular deep-learning model in R that keeps widening what counts as a tabular task.
◆ Current state
tabnet ports the TabNet attentive tabular architecture to R on torch, wired into tidymodels through parsnip so it slots into workflows, tuning, and case weights like any other engine. The model surface has grown well past plain supervised fitting: unsupervised pretraining, missing values in predictors, multi-outcome fitting, hierarchical multi-label classification, and built-in explainability via tabnet_explain(). The 0.9.x line has been consolidating rather than adding, with 0.9.0 finally making hierarchical classification work correctly by accounting for the ancestor matrix.
◆ Where it's heading
Two threads run through the release history. The first is task surface — each minor version tends to admit a class of problem the model previously could not express, from missing data to hierarchy to imbalanced binary outcomes. The second is torch-level performance and correctness, visible in the torch_ignite_adam default that cut pretraining time roughly 30% and the fix for optimizers frozen after checkpointing on cuda and mps. Tidymodels integration is treated as a first-class obligation, with parsnip breaking changes tracked release by release.
◆ Prediction
The hierarchical path is the least finished: 0.5.0 introduced it and 0.9.0 only just made it effective, so the next releases most likely extend evaluation and explainability to hierarchical fits rather than adding another task type.
Alternatives to lime and tabnet
Other Analytics products tracked by Sparkpulse, ranked by recent ship velocity. Each card links to a full editorial trajectory and lets you pivot into a head-to-head comparison with either lime or tabnet.
DoseFinding
New stewardship at openpharma, then two releases adding the methods MCP-Mod was missing
webmockr
The stubbing library added httr2 support, then spent a year cutting itself free of everything else
crul
crul took mocking back from webmockr and made it a property of the client itself
dendroNetwork
Six releases, six identical bodies — the feed carries the package abstract instead of release notes
chattr
chattr deleted every LLM integration it had written and outsourced the lot to ellmer
Recent activity from lime and tabnet
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
- 21d agotabnetvip dependency moves to r-universe
- 2mo agotabnetHierarchical classification made effective, augment() added
- 6mo agotabnetentmax15 and sparsemax15 masks, AUM loss for imbalanced data
- 8mo agolimexgboost compatibility restored across versions
- 1y agotabnetBugfix release for R 4.5 and dials tuning
- 2y agotabnetCase weights and warm-start parameters via parsnip
- 2y agotabnetHierarchical multi-label classification via data.tree ⚡
- 3y agolimeMaintainer handover and general upkeep
- 5y agolimeShiny dependencies moved to Suggests
- 6y agolimeNamespace fix for glmnet changes
- 7y agolimegower_pow added and lambda aligned with the Python implementation
- 8y agolimeh2o support, NA handling, and date columns held during permutation
Frequently asked questions
What is the difference between lime and tabnet?
Both compete on the same themes — explainability — within Analytics. tabnet is currently shipping more aggressively (velocity 2.5 vs 0.0), with 0 editorial sparks in the last 30 days against 0. See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Is lime better than tabnet?
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. tabnet is currently shipping more aggressively (velocity 2.5 vs 0.0), with 0 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other Analytics products to evaluate alongside.
What are the best alternatives to lime?
Top lime alternatives in Analytics are ranked by recent ship velocity. Browse the "lime alternatives" section above for the current picks, or visit /alternatives/lime for the full list with editorial commentary on each.
What are the best alternatives to tabnet?
Top tabnet alternatives in Analytics are ranked by recent ship velocity. Browse the "tabnet alternatives" section above for the current picks, or visit /alternatives/tabnet for the full list with editorial commentary on each.