LightTAR: target-aware dimensionality reduction for multimodal tabular learning
Technion · hf · 2026-11-19
Researchers at Technion propose LightTAR for multimodal tabular learning, where standard PCA reduction keeps input variance but may drop target-relevant information. LightTAR augments PCA features with a few target-aware features extracted via regularized partial least squares from the final and an intermediate encoder layer, adapting dimensionality reduction to the prediction task and improving MMTL performance with minimal runtime overhead.
More from Research
- Retrieval eval insight: models topping BEIR/DL19/DL20/MIRACL are likely below SOTA — bclavie · 2026-11-19
- Sandia Labs' VASt model predicts high-entropy alloy properties from tens of samples — CurieuxExplorer · 2026-11-19
- FoodTruck Bench, an LLM agent business-sim benchmark, lands on Papers with Code — NielsRogge · 2026-11-19
- MIT reads and steers LLMs' internal physics intuition via hidden-state shifts in Gemma — ProfBuehlerMIT · 2026-11-19
- harness-hub open-sources RL-trained LoRA adapters tuned per agent harness, recipe coming Friday — yb2698 · 2026-11-19
- Freenet releases new video on network routing with old-school ML — Sanity · 2026-11-19