Using an LLM to Derive Interpretable Labels for PLSR Components in Drug Effect Space
Josikinz · x · 2026-10-03
- The author's team validated discriminating drugs from experience-report embeddings, then used PLSR to map covariance between biochemistry and subjective experience, positioning drugs in a multidimensional effect space.
- To derive interpretable labels for each PLSR component, an LLM synthesized grab bags of experience themes while a blinded human expert (the author) performed the same task in parallel — a neat LLM-plus-human validation workflow.
- More discussion of LLM-based summarization is linked in the thread.
Related event: NLP and LLMs Map Psychedelic Subjective Experiences(4 posts)→
More from Research
- Researchers call for simplest-possible baselines in ML×Bio papers over flashy ablations — anshulkundaje · 2026-10-03
- Meta's Muse Spark helps disprove an evolution algebras conjecture, one of six AI-math papers — AIatMeta · 2026-10-03
- Meta partners with mathematicians on six papers, solving five open research problems with Muse Spark — AIatMeta · 2026-10-03
- PDFs often lie about or omit word and character positions — 5 failure modes explained — VikParuchuri · 2026-10-03
- NBER paper: technical change 'commoditizes' labor, boosting productivity but suppressing wages — PeterHndrsn · 2026-10-03
- KernelBench-Verified: no frontier model beats PyTorch when evals get strict, Meta/Stanford find — lmoroney · 2026-10-03