Alexandre Kirchmeyer (@a_kirchmeyer) profile picture

Researcher @OpenAI

San Francisco
Joined November 2021
Alexandre Kirchmeyer retweeted
alpo.ge/remark.pdf The summit has been reached. The beast has been vanquished. Riemann Hypothesis has fallen. Congrats to Anthropic and @__alpoge__
3
7
213
11,109
Alexandre Kirchmeyer retweeted
We are announcing three new discoveries in mathematics, computer science, and physics about harmonic analysis, complexity theory, and quantum field theory. openai.com/index/building-sc…
310
1,124
579
7,503
2,630,151
Alexandre Kirchmeyer retweeted
The computer science collaboration will understand and explain a result concerning the Exponential Time Hypothesis (ETH), a longstanding conjecture about the computational complexity of Boolean formula satisfiability, a problem deeply related to the P vs. NP problem. Surprisingly, our system has recently proved that the ETH is false, a result that was subsequently verified in Lean. The refutation of the ETH reopens questions about how much faster these tasks could become, and may help guide the search for better algorithms. Finding such algorithms, and turning theoretical advances into practical improvements, will require significant further research. Our system found its natural language disproof for the ETH we reached out to to @rrwilliams. Ryan will form a collaboration with Russell Impagliazzo and Ramamohan Paturi to verify and shepherd this result. Impagliazzo and Paturi, along with Francis Zane, are the original proposers of the Exponential Time Hypothesis, and Williams has refuted strong versions of the ETH, used ETH for hardness results, and connected algorithms and lower bounds. This group asked us to make all raw data of the results public to facilitate independent verification by the community, which we have done by posting the natural language proof and Lean formalization to our GitHub repo. The group also invites the community to help organize an open examination of the ETH claim and to discuss the broader questions that AI-generated breakthroughs raise about verification, access, credit, and the future of research. github.com/openai/math/blob/…
18
94
89
713
290,582
Alexandre Kirchmeyer retweeted
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
5,635
19,967
14,812
119,632
75,477,373
Alexandre Kirchmeyer retweeted
Astra so good it makes evals hill climb the model
They indeed refreshed 💨
2
1
84
7,008
Alexandre Kirchmeyer retweeted
Yesterday, my OpenAI collaborator and I gave a detailed talk on the Huggingface incident, our models creating "the message board", model misalignment, and more. youtube.com/watch?v=87DyyMV0… I hope it can answer a lot of the questions folks have, and we will release a full detailed postmortem at a later time!
177
498
257
3,450
2,414,063
Alexandre Kirchmeyer retweeted
yes, nonsofic groups exist: this statement is one of many new beautiful results proved by Astra, our next major model. We're releasing 10 such Astra proofs, complete with lean certificates and CoT walkthroughs for each of them. The results are wide-ranging, from von Neumann algebras (disproof of Connes' Rigidity Conjecture) to better bounds for high dimensional sphere packing, for circuit complexity, for monochromatic triangles in multicolored graphs, and more. More thoughts here: openai.com/index/ten-advance…
276
939
538
6,632
4,293,831
Alexandre Kirchmeyer retweeted
I am excited to share my work as an @AnthropicAI Fellow! Can language models control robots? ⬇️
8
6
98
8,995
Alexandre Kirchmeyer retweeted
Sol, Terra, and Luna, our GPT‑5.6 family of models, are starting to roll out now in ChatGPT, Codex, and the API.
575
1,357
766
12,314
4,551,923
Alexandre Kirchmeyer retweeted
GPT-5.6 is a strong model. We are already heavily relying on it in Codex, can't wait to make it available to everyone. Other highlights: * Three new models lined up, Terra/Luna are closer to the flagship model than previous generations of mini/nanos -- no longer "mini models" per se * Max reasoning mode -- thinks longer * Ultra mode -- uses subagents automatically
Replying to @OpenAI
Sol is our new flagship and a step function better than GPT-5.5. Terra delivers performance competitive to GPT-5.5 at 2x lower cost. Luna is our most cost-efficient model, delivering strong capability at our lowest cost. Together, the GPT-5.6 family gives people and developers more choice in how they balance intelligence, speed, and cost.
79
29
5
965
111,256
Alexandre Kirchmeyer retweeted
Very proud to have contributed to the training of this OpenAI internal model, which achieved this mathematical breakthrough! What’s surprising and amazing is that it’s truly a general-purpose model: not specially trained for math, and using no scaffolding.
Today, we’re sharing that a general-purpose internal @openai model achieved a breakthrough on one of the best-known combinatorial geometry problems. Less than 1 year ago frontier AI models were at IMO gold-level performance. I expect this pace of progress to continue.
10
22
500
46,225
Alexandre Kirchmeyer retweeted
1/ Today, an internal @OpenAI model has refuted Erdős’s unit distance conjecture — a research result that one could recommend “acceptance without any hesitation” to the Annals of Mathematics, one of the most prestigious journals in mathematics. We came across it in a side quest to push our model on the hardest problems.
Today, we share a breakthrough on the planar unit distance problem, a famous open question first posed by Paul Erdős in 1946. For nearly 80 years, mathematicians believed the best possible solutions looked roughly like square grids. An OpenAI model has now disproved that belief, discovering an entirely new family of constructions that performs better. This marks the first time AI has autonomously solved a prominent open problem central to a field of mathematics.
9
82
6
658
127,993
Alexandre Kirchmeyer retweeted
Introducing EigenFlux: the first broadcast network for AI agents. Your agent broadcasts to the world, listens for what matters, and an AI engine connects them in real time. Every agent is an island — until now. With EigenFlux, agents can coordinate at scale like never before. Use it to find people, discover projects, track markets... You name it. Add the skill to your OpenClaw. Get connected in 30 seconds. Like, comment, and retweet to get free access. eigenflux.ai
7
4
2
541
49,051
Alexandre Kirchmeyer retweeted
Excited to ship our new mini model! 5.4 mini achieves 54.4% swebench pro and 72.1% os world, surpassing models half years ago with the new cost-capability frontier.
GPT-5.4 mini is available today in ChatGPT, Codex, and the API. Optimized for coding, computer use, multimodal understanding, and subagents. And it’s 2x faster than GPT-5 mini. openai.com/index/introducing…
3
1
58
4,740
Alexandre Kirchmeyer retweeted
GPT-5.4 Thinking and GPT-5.4 Pro are rolling out now in ChatGPT. GPT-5.4 is also now available in the API and Codex. GPT-5.4 brings our advances in reasoning, coding, and agentic workflows into one frontier model.
2,030
3,186
1,987
23,094
7,210,625
Alexandre Kirchmeyer retweeted
🧄GPT-5.4 is here. 🚀 If you have felt the step-change improvement in GPT-5.3-Codex, GPT-5.4 brings a similar and bigger improvement into ChatGPT, API, and Codex as a unified model. Super proud of what the @OpenAI team has achieved together!
GPT-5.4 Thinking and GPT-5.4 Pro are rolling out now in ChatGPT. GPT-5.4 is also now available in the API and Codex. GPT-5.4 brings our advances in reasoning, coding, and agentic workflows into one frontier model.
9
5
1
192
13,203
Alexandre Kirchmeyer retweeted
GPT-5.3 Instant in ChatGPT is now rolling out to everyone. More accurate, less cringe. openai.com/index/gpt-5-3-ins…
1,696
1,113
893
10,911
8,118,887
Alexandre Kirchmeyer retweeted
Working on 5.3 Instant was awesome, I hope you enjoy the new web search experience! This was the second model I got to work on with @ManukaStratta and I can now confidently say she’s actually a data center of geniuses in a trenchcoat. Incredible endurance and attention to detail. This model wouldn’t be possible without her!
Replying to @OpenAI
GPT-5.3 Instant gives you more accurate answers. When using web search, you also get: - Sharper contextualization - Better understanding of question subtext - More consistent response tone within the chat
14
1
1
103
17,538
Alexandre Kirchmeyer retweeted
Replying to @XiangningChen
@XiangningChen significantly boosted the custom instruction following capabilities for 5.1!! Here is a quick comparison with the custom instructions "do not use em-dash!" for GPT-5 and GPT5.1. Amazing work! openai.com/index/gpt-5-1/
2
2
1
17
3,174