Intology's Locus automates molecular design, beating expert solutions and LLM agents
ChengleiSi · x · 2026-11-03
Intology unveiled Locus, an automated research system for molecular design that claims to outperform SOTA expert solutions and LLM-agent baselines. The key idea: autonomously leveraging proprietary data, addressing the common AI-for-science pain point of out-of-distribution tasks, limited data, and expensive noisy evaluators.
Benchmarks cover three real-world settings: training molecular property prediction models on a proprietary task, fine-tuning Boltz-2 for small-molecule potency prediction, and building antibody design workflows around BoltzGen, Boltz-2, ESMC, and ESMFold2. Chart comparisons show Locus leading Codex and Claude Code baselines across property prediction, antibody design (ipSAE), and potency prediction (0.6 test correlation).
More from Research
- ParallACT benchmark shows multi-agent communication barely helps: 16%→32% at 2.23x token cost — EliasEskin · 2026-11-04
- Multi-agent messaging barely helps: 61% of agreed runs still hit merge conflicts — EliasEskin · 2026-11-04
- Parallel agents collapse: Opus joint success drops from 62% to 16% — EliasEskin · 2026-11-04
- "Deep lifts" paper: ReLU networks adaptively generalize even with the dumbest random optimizer — yuxiangw_cs · 2026-11-04
- DR Pixel: open-source page-image deep research agent, 27B model matches GPT-5.4 — sewon__min · 2026-11-04
- FlexGlue: open-source pixel-perfect image matching that rejects doppelganger pairs — ducha_aiki · 2026-11-04