Robot sorting eval: Claude Fable 5.2 finishes 8720, GPT-6 Astra lands only 8 and 7
qinzytech · x · 2026-11-10
Embodied Analysis compared frontier models on a robot sorting task: Claude Fable 5.2 completed it, finishing 8720, while its index leader GPT-6 Astra ended with only 8 and 7 correctly placed. Both models placed three digits initially, but Astra's 2 drifted off its pad — a caution that later actions can undo earlier progress. Context: Anthropic just launched Claude Fable 5.2 and Mythos 5.2, touting coding, reasoning, and vision gains.
More from Embodied
- NEO-VLA: encoder-free native VLA turns pixels into actions, beats encoder-based baselines — liuziwei7 · 2026-11-11
- Agent tunes a live camera for ML training: focus lock 190, white balance 4600K — TheZachMueller · 2026-11-11
- Stanford's BEHAVIOR Challenge: full-task success jumps 12% to 48% in one year, Huawei team wins — drfeifei · 2026-11-11
- CoRL 2026 slide trains an LSTM in 1 minute to predict robot joint torque — shaohua0116 · 2026-11-11
- Musk claims Cybercab will do majority of passenger miles; commenters cite unmet demand — RachelVT42 · 2026-11-11
- Newton: An open-source GPU-accelerated physics engine built on NVIDIA Warp — ChongZzZhang · 2026-11-11