Inside Terminal-Bench-Science: Defining the Next Gen of Science Agents
ajratner · x · 2026-08-28
This post details Terminal-Bench-Science v0.1, aiming to define the next generation of scientific coding agents, akin to a 'Claude Code for Science.' The benchmark features realistic computational workflows across physical, life, and mathematical sciences. Beyond the difficulty leaderboard, TB-Science sets a high bar for benchmark construction with multi-stage expert review, requiring deep domain expertise to build.
Related event: Stanford-led Terminal-Bench-Science Debuts; Claude Opus 5 Passes Just 30%(19 posts)→
More from Research
- Pathway CSO pitches BDH as post-Transformer architecture uniting memory, reasoning, adaptation — schwentker · 2026-10-15
- EmbodiedEvo tops RoboDojo at 38.95% success by letting robots inherit failed practice — thetripathi58 · 2026-10-15
- 797GB of n-gram counts in CPU RAM let 1B models match 57% more compute — PangWeiKoh · 2026-10-15
- Google's HARNESS² evolves agent harnesses without verifiers, up to +11.5 points on WorkBuddy-Bench — dair_ai · 2026-10-15
- Subbarao Kambhampati to dissect the 'mythos of LLM thinking tokens' at UMN — rao2z · 2026-10-15
- How 3D Gaussian Splatting gives robots spatial memory of what's out of view — TheTuringPost · 2026-10-15