Metal TreeSHAP claims 85.1x GPU speedup for exact TreeSHAP on 64,000 rows
bingxu_ · x · 2026-07-22
Tabul AI says it is sharing Metal TreeSHAP, a GPU-accelerated version of TreeSHAP for model interpretability.
The chart highlights a benchmark on Auto MPG with 400 trees, claiming exact TreeSHAP and showing a 85.1x GPU speedup at 64,000 rows.
The post frames Shapley values as one of the most widely used tools in ML interpretability, and positions the new implementation as a major performance improvement for larger datasets.
Related event: Tabul AI Releases Metal TreeSHAP to Accelerate Model Interpretability(2 posts)→
More from Research
- Paper: data schedules and learning rate schedules should be designed together — alifcoder · 2026-10-27
- Health benchmark study: quality and diversity of eval items beat quantity — keunwoochoi · 2026-10-27
- Across 7 tasks, highest reasoning setting isn't the best for most open models — keunwoochoi · 2026-10-27
- Open Models Match GPT-6 on MedMCQA but Trail 17 Points on Hard Clinical Cases — keunwoochoi · 2026-10-27
- Health-Focused Benchmark Reranks LLMs: Muse Spark Open Jumps from #9 to #1 — keunwoochoi · 2026-10-27
- Open-source Latch compiles PyTorch into training megakernels, 2.75x faster on H100 — sguada · 2026-10-27