Currently looking for full-time RS/RE opportunities at frontier AI labs. Available immediately and open to relocating to SF.
https://t.co/pXivmbRwST
Would love to connect!❤️
📣 Our team at AWS is hiring summer 2027 interns to work across pre-training, mid-training, and post-training, including alignment.
Interested? DM me! 📩
RTs appreciated! 🙏
@l32zhang@GoogleDeepMind Hi Lei! I’m Jiaqi Wang, a final-stage PhD student at CUHK working on multi-turn agent RL. I’ve also worked on LLM post-training at Tongyi Qwen and MiniMax. I just applied for the role and would love to chat briefly about what your team is working on if you have time.
🚨 Jane Street vient de sortir son agent tuteur
Et c'est un monstre d'apprentissage
Et ils l'ont rendu public, ils viennent de mettre un skill IA sur GitHub, licence MIT
En gros, c'est un tuteur socratique. Il vous pose des questions pour vous forcer à raisonner au lieu de lâcher les réponses déjà toutes cuites
Concrètement, ça fait quoi
> Diagnostic de ce que vous savez déjà
> Carte de compétences, niveaux 0 à 5
> Une seule question à la fois
> Vous apprend le sujet de A à Z
> Vous permet de creuser un sujet ou d'en découvrir d'autres
> Révisions espacées + contrôle final
Comment l'utiliser : 3 options
1/ Sur GrokBot
Directement commenter sous le post @bot "Crée un bot et charge le skill".
2/ Sur Claude
Customize → Skills → Upload a skill
Zipper le dossier socratic-tutor (avec le SKILL.md dedans), l'uploader et l'activer
Puis utiliser "Teach me..." ou /socratic-tutor
3/ Sur ChatGPT
Créer un Project, coller le contenu du SKILL.md dans les instructions du projet
Chaque chat du projet démarre en mode tuteur
Vous pouvez aussi demander des illustrations à chaque usage pour le rendre plus propre
https://t.co/nNyN9TykF0
I am at @PKU1898 for a AI4Math seminar. And there are more people than the seats can hold.
Two takes:
(1) What is the most important in the AI Ara?
The answer is: HEALTH!
(I usually forget this answer since most of the time, the answer I get is "human interaction", but "health" is the ultimate answer that we constantly forgets.
(2) What should we do in the future when AI become so powerful?
We should do what is valuable rather than what is innnovative, and these two are not the same.
Announcement:
I have indeed been diagnosed with the disease that everyone fears, but it was caught at a very early stage.
I have already scheduled my surgery, and I thank everyone for your concern.
Honestly, I no longer care about so-called money or prestige. I just want to do what I can within my academic field and work on AI things that bring me joy (the doctor says this might also help with my recovery).
I have let go of all past interpersonal entanglements, as well as others' prejudices and discriminations against me.
In the face of life and death, these things are so insignificant. It is this illness that has made me rethink the meaning of life and death: one should not live for the approval of others, but persist for the things we love and the people who love us.
I will continue to share because it is a meaningful thing to do.
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: https://t.co/bOWCcFhCVq