Xaira unveils X-Scientist: autonomous research agent beats Codex by 7.4 points on DrugDiscoveryBench
BoWang87 · x · 2026-10-22
Xaira introduces X-Scientist, an agentic AI platform for drug discovery aimed at recursive self-improvement: it designs experiments, builds models, evaluates results, and decides what to investigate next, with asynchronous model–harness co-evolution as the key mechanism. In a PerturbQA pilot (4-hour allowance, $100 API cap), X-Scientist reached the best validation AUROC of 0.6789 vs AutoScientists' 0.6728 and AutoResearch's 0.6441, while spending only $30.52 in API costs versus $93.27 for AutoScientists.
More from coding & agent
- Classic 'too many tabs' meme applied to coding agents: send help — mrdrozdov · 2026-10-22
- YC F26 startup Latent reads model internals to catch hallucinations at 530x lower cost — ycombinator · 2026-10-22
- We need an L0.5 for continual learning: agent state stacks up in 4 layers — willcb · 2026-10-22
- DuckDB + Hugging Face Amsterdam workshop: build an AI feedback pipeline over 2.1M Amazon reviews — lhoestq · 2026-10-22
- Veracode report: AI now writes half of committed code but only 56% pass security checks — WeldPond · 2026-10-22
- Will Brown: Continual Learning in the Harness Is Like a Computer Without RAM — willcb · 2026-10-22