AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

The Illusion of Generalization in Tabular Language Models

Aditya Gorla, Ratish Puduppully

ICML 2026 · Large language models › Language, safety and interpretability · approach: Controlled text generation · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add general instruction tuning (Alpaca) to Base Llama (ablation)Llama-3-8B · UniPredict classification (119 datasets)accuracy (%): 47.7 -> 58.6Improved
add tabular pretraining on T4 on top of instruction tuning (Tabula-8B vs Alpaca) (ablation)Alpaca · UniPredict classification (119 datasets)accuracy (%): 58.6 -> 63.5Improved
exclude 5 contaminated datasets from the decomposition (ablation)instruction-tuning contribution share · UniPredict classificationshare of gain (%): 69% -> 73%Mixed
add quartile-format examples to Alpaca (Alpaca+Q) (ablation)Alpaca · UniPredict quartile classificationlift (pp): 5.6 -> 25.8Improved
add tabular pretraining (Tabula-8B vs Alpaca+Q) on non-stock quartile datasets (ablation)Alpaca+Q · UniPredict quartile non-stocklift (pp): 6.7 -> 6.5No clear effect
add tabular pretraining (Tabula-8B vs Alpaca+Q) on stock/financial quartile datasets (ablation)Alpaca+Q · UniPredict quartile stock/financiallift (pp): 48.2 -> 66.1Improved
report lift over majority baseline instead of raw accuracyTabula-8B · Stroke Predictionaccuracy vs baseline: 0.949 accuracy vs 0.949 majorityHurt
evaluate on binary instead of quartile tasksTabula-8B · UniPredictmedian lift (pp): 32.9 -> 2.5Hurt
evaluate on categorical instead of quartile tasksTabula-8B · UniPredictmedian lift (pp): 32.9 -> -0.3Hurt
use Base Llama without instruction tuningLlama-3-8B · UniPredict classificationdatasets at or below majority baseline (%): 60.5 -> 75.6Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy