Masking giveaway terms barely dents LitGene's gene-relation scores, study shows
bravo_abad · x · 2026-10-21
To test whether an AI gene model learned biology or just word matching, strip the giveaway terms. Macaulay et al. evaluated LitGene, which uses contrastive learning to link gene descriptions to known functions. Removing function names and synonyms altered 84.5% of gene summaries yet barely changed the model's ability to separate related from unrelated genes; removing pathway names dropped the retrieval score only from 0.5510 to 0.5493—evidence it learned relationships beyond surface cues.
More from Research
- A real larva's connectome — 2,675 cells, 14,066 connections — now swims in a 3D simulation — eptwts · 2026-10-21
- Stanford's T2-z lets robots probe hidden physics with a poke, lifting success from 34% to 73% — stanfordnlp · 2026-10-21
- Fei-Fei Li team unveils FACT, a force-aware foundation model for robot dexterity — stanfordnlp · 2026-10-21
- Upcoming paper claims big models can handle latency-critical real-time robotics — ManlingLi_ · 2026-10-21
- NVIDIA open-sources Flora, a robotics data engine 125x cheaper than GPT-6 Astra — DrJimFan · 2026-10-21
- openRxiv explains its approach to AI amid growing researcher adoption — anshulkundaje · 2026-10-21