GuideLoom: first open-source dataloader for training interpretable LLMs at scale
juliusadml · x · 2026-10-28
GuideLabs open-sourced GuideLoom, billed as the first dataloader for training interpretable language models at scale. The team has used it to train models from 100M to 32B params.
It stores data as documents, token streams, or token streams with span-level annotations; during training a PyTorch DataLoader shuffles across the entire dataset, lets every GPU read its own samples in parallel, and resumes from the exact batch after a restart. Install: pip install guideloom[torch].
More from Research
- Researcher predicts molecular simulation will accurately model most chemistry and biology within years — TimothyDuignan · 2026-10-28
- Debate over AI vs. human text: paper doesn't prove detection is technically unsolvable — _akpiper · 2026-10-28
- Coarse-graining can scale all-atom sims orders of magnitude, researcher argues — TimothyDuignan · 2026-10-28
- Are agent harnesses dying? Distillation experiments move scaffolding into the model — rseroter · 2026-10-28
- Abstraction neurons as embeddings beat SOTA on word similarity, fail badly on sentence retrieval — mayfer · 2026-10-28
- Purdue's IDEAS Lab demos GPS-free drone that tracks runners through dense forests on-chip — DavidSKrueger · 2026-10-28