Загрузка видео...

Не удалось загрузить видео

На главную

Astra began to saturate a computer-use eval we haven’t even released yet. Astra was not only able to kill several hostile mobs in Minecraft, but it made it all the way to building a Nether portal in less than three hours. We believe that if we ran it longer,...

30,212 просмотров • 6 дней назад •via X (Twitter)

Комментарии: 16

Фото профиля Vals AI
Vals AI6 дней назад

Astra killing zombies

Фото профиля Vals AI
Vals AI6 дней назад

Astra does the trap door trick to mine more efficiently

Фото профиля Vals AI
Vals AI6 дней назад

Astra boats back to land

Фото профиля Vals AI
Vals AI6 дней назад

Astra enabled “Keep Inventory,” allowing it to progress faster. However, its ability to operate in Minecraft represents a huge leap over previous models.

Фото профиля imjustnewatai
imjustnewatai6 дней назад

Can’t you make it where it pauses during its thinking? I feel like it would make it a lot better.

Фото профиля kiri49
kiri495 дней назад

if an open weights model is ever able to do this anytime soon, that would be very impressive. but this is otherwise very cool for an ai to actually be able to play minecraft like this, very impressive showing from openai

Фото профиля aizk ✡️
aizk ✡️6 дней назад

Astra building a 3x3 door was my AGI moment

Фото профиля John Rood
John Rood6 дней назад

the next benchmark has to randomize the world on every run. a fixed three-hour Minecraft trajectory becomes a demo fast. change the spawn, inventory, goals, and hazards to see whether the agent learned the game or one path through it.

Фото профиля Geoff Kwitko
Geoff Kwitko5 дней назад

Impressive

Фото профиля Jesse Pan
Jesse Pan5 дней назад

MCSpeedrunBench could be interesting

Фото профиля doof
doof5 дней назад

It's over if Astra gets on Cerebras.

Фото профиля ☃️Darth thromBOOzyt📯
☃️Darth thromBOOzyt📯5 дней назад

Wow. Next step might be something like the recent open world Zelda games. Wondering if it can do these too (or challenge runs that require deep planning ahead)

Фото профиля 刘朝 Zhao Liu
刘朝 Zhao Liu6 дней назад

The long-horizon trace is the real benchmark: survival after deaths, inventory state, recovery from bad plans, and the number of human interventions. Publishing the harness and full replay would turn a three-hour demo into a comparable computer-use eval.

Фото профиля Naren
Naren6 дней назад

Release it! this way we can compare it with other models

Фото профиля anytan
anytan6 дней назад

nether portal in under 3 hrs.

Фото профиля Lucas
Lucas5 дней назад

more tokens. Beating the game would make history

Похожие видео