Autoresearch agent eval: hypothesis-distinguishing tests up from 16.8% to 28.0%
anirudhg9119 · x · 2026-10-14
The core takeaway: scaling autoresearch is about learning which investigation deserves the next unit of compute, and assigning credit by its long-term consequences.
In an LLM-judged study of 400 matched trajectory pairs, the learned policy:
- Raised hypothesis-distinguishing tests from 16.8% to 28.0%
- Cut repeated minor tweaks to stalled ideas from 33.5% to 18.8%
The policy investigates unresolved uncertainty more often instead of spinning in place.
More from Research
- China's GigaAI Opens First Provincial-Level General World Model Lab in Beijing — pstAsiatech · 2026-10-14
- Mosaic model engineers cell states to expose hidden disease variants, opening chromatin at 28% of AMD loci vs 5-7% at rest — anshulkundaje · 2026-10-14
- NeurIPS paper: visual in-context learning lifts VLM accuracy by up to 27% — cindy_x_wu · 2026-10-14
- Pythia team releases protein-RNA binding atlas preprint probing why the model predicts — anshulkundaje · 2026-10-14
- pqnld: A Rust Service That Turns Your Existing LLM Into a Closed-Set Decision Model via Logprobs — Difficult_Ad_2778 · 2026-10-14
- Mixed-precision, mixed-accelerator samplers cut RL training cost with off-policy correction — xiuyu_l · 2026-10-14