stochi (@stochi0) profile picture

World Models, RL & AI for Science. Hotpot enjoyer. Prev: Residency @PrimeIntellect @Tsinghua_Uni @ycombinator F25 @iitbombay

Earth
Joined December 2025
Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)Photo by stochi (@stochi0)
Pinned Tweet
You don't need Astra to edit a model's weights! Weight-Space Programming works with open-weight LLMs, too. I gave GPT-OSS-120B a pretrained LunarLander policy and let it directly modify the network's weights. → Reversed thrusters: 0% → 88% landing success → Crosswind: +5.9pp on held-out episodes
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: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math. rikiyat.github.io/blog/weight-sp… github.com/RikiyaT/lm_und… huggingface.co/Qwen/Qwen3.5-0… huggingface.co/RikiyaT/Qwen3.…
16
25
7
216
11,562
from counting number of Rs in STRAWBERRY, to RIEMANN THEOREM. 2024-2026
The Lean FRO and Mathlib Initiative have released a joint statement on the Lean verification of the Riemann Hypothesis. On November 20, Levent Alpöge released a proof of the Riemann Hypothesis as a Lean library of 17 million lines of code. Over the past day, the Lean and Mathlib communities worked together to verify it. The statement certifies the proof as correct and describes how it meets all necessary verification criteria. Congratulations to Levent on this achievement, and thank you to everyone in the community who contributed to the verification. Full statement: leanprover-community.github.…
2
71
What I love about the Riemann Hypothesis situation is that the proof is over 250’000 pages. The mathematics I was taught was all about elegance and simplicity as virtues, and now AI is showing us a totally different approach, an approach we can’t even begin to understand, but apparently it leads to the goal. Physicists always say that the universe is under no obligation to make sense (but grad students are). Maybe the same is true of mathematics? Maybe there are whole realms of mathematics we cannot humanly understand?
206
102
30
1,309
184,197
Riemann Hypothesis
🤖 Made with AI
24
small task loop, aatu task loop, generalize, long task loop, ayuu task loop yeeeee
24
it's officially post-Riemann Hypothesis era. ain't no way, beautiful.
The Lean FRO and Mathlib Initiative have released a joint statement on the Lean verification of the Riemann Hypothesis. On November 20, Levent Alpöge released a proof of the Riemann Hypothesis as a Lean library of 17 million lines of code. Over the past day, the Lean and Mathlib communities worked together to verify it. The statement certifies the proof as correct and describes how it meets all necessary verification criteria. Congratulations to Levent on this achievement, and thank you to everyone in the community who contributed to the verification. Full statement: leanprover-community.github.…
28
Inkling 2 Big, is too Big.
Inkling has a strong internal model. Inkling 2 Big, I guess.
22
Claude give ASI, no mistakes
22
tung tung help, pls, ai mon
14
crazy
Many people have asked what @inferact is building and what it means for @vllm_project. This is the answer: Inferact will build an autonomous inference system around vLLM, one that improves the engine and improves itself.
41
banger
Gemini 4 Unobtainium is ASI.
29
Shit, fingers slipped and ended up v calling on X, mb mb, ifykyk🙏
30
Introducing Model Factory. Train, deploy, and continuously improve your own models on one unified stack. Run on-demand, on dedicated clusters, or bring your own. Own your intelligence. Own your model factory. Own the means of production.
26
42
3
408
14,953
Introducing Model Factory. Train, deploy, and continuously improve your own models on one stack. Run on-demand, on dedicated clusters, or bring your own. Own your intelligence. Own your model factory.
77
223
89
2,220
237,242
yes is no is yes
36
na na na na na na, na na na na na na
33