DSAIEngineering

DSAIEngineering

June 2026

[P32] Beta distribution
An introduction to the distribution and its applications.
 • Mohit Saharan
[P31] Architecture of TabICLv2: quantile predictions for regression
How TabICLv2 models regression with 999 conditional quantiles, pinball loss, and distribution reconstruction.
 • Mohit Saharan
[P30] Architecture of TabICLv2: many-class classification
How TabICLv2 handles classification tasks with more than 10 classes by decomposing labels on the embedding side and predictions on the output side.
 • Mohit Saharan
[P29] Architecture of TabICLv2: query-aware scalable softmax
How TabICLv2 uses query-aware scalable softmax (QASSMax) to prevent attention fading in tabular in-context learning.
 • Mohit Saharan
[P28] Architecture of TabICLv2: compression-then-ICL
How TabICLv2 compresses target-aware feature tokens into row representations, then uses dataset-wise in-context learning to predict test rows.
 • Mohit Saharan
[P27] Architecture of TabICLv2: target-aware embedding
How TabICLv2 uses target-aware embedding to add training labels to tabular in-context learning tokens while preventing label leakage in test rows.
 • Mohit Saharan
[P26] Architecture of TabICLv2: repeated feature grouping
A technical guide to TabICLv2 repeated feature grouping: why similar columns confuse encoders, how circular shifts add context, with NanoTabICL…
 • Mohit Saharan

May 2026

© 2026 Mohit Saharan · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture