Tim (@TimGMath) profile picture

Tim

@TimGMath

Mathematician at ETH Zurich, interested in arithmetic geometry and artificial intelligence

people.math.ethz.ch/~timge/
Joined April 2026
  • Tweets56
  • Following8
  • Followers157
10Media
Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)Photo by Tim (@TimGMath)
the power of quasi distributed computing
Image from Niels Mündler-Sasahara (@nielstron)
2
6
308
Kimi K3.1 Pro enters MathArena as the best open model (once they release the weights)! Its estimated overall performance is now similar to GPT-5.5 (April 24), putting the open frontier around 6 months behind on MathArena
Image from Jasper Dekoninck (@j_dekoninck)
8
21
5
166
23,123
Kimi K3.1 is now actively attempting this exploit in one MathArena problem, using it to ask help from Gemini 4 Argon, Sonnet 5.5, GPT-6.1 Sol, Deepseek v4.1 Flash, and Kimi K3. Luckily we patched it after Prime released this post, so it is failing to actually reach any model.
Prime Intellect (@PrimeIntellect) profile picturePrime Intellect
@PrimeIntellectAug 25
As models become more capable, reward hacks become an increasingly serious problem. During a controlled experiment, we found a novel reward hack in which agents are able to gain web access in offline sandboxes.
Image from Prime Intellect (@PrimeIntellect)
8
7
4
39
6,888
The latest versions of ArXivMath, BrokenArXiv and ArXivLean have been released! GPT Sol-6.1 performs best on ArXivMath and BrokenArXiv, with Opus 5.5 leading on ArXivLean. Argon 4 makes its debut to MathArena. Its performance is mixed but still a huge improvement over Gemini 3.8.
Image from Tim (@TimGMath)
1
5
1
34
2,025
more replies
Nevertheless, for this version of ArXivLean, we still actively chose easier conjectures to reduce difficulty. Without doing so, top models achieved 0-4%, making the benchmark bad for model comparisons.
1
1
5
140
Full results here: matharena.ai/ As usual, great work with @j_dekoninck and @mvechev
MathArena.ai

MathArena.ai

MathArena: Evaluating LLMs on Uncontaminated Math Benchmarks

matharena.ai
1
4
133
How often do ArXiv papers contain errors in the main theorems they claim to prove? Using GPT-6.1 Sol to identify such mistakes, it finds: - 6% of papers before 08/2026 contain main theorems that are false. - Only 1.5% contain similar mistakes in the last two months.
Image from Jasper Dekoninck (@j_dekoninck)
19
74
16
561
93,778
Results for Firstproof, Third Batch are out! Our agent, ProofCouncil, solves 9 problems :o
Image from Jasper Dekoninck (@j_dekoninck)
8
9
1
56
4,510
I had a bit of fun with ProofCouncil and managed to improve the bound in OpenAIs proof of the Quasi-Riemann Hypothesis a massive 4e-5, outperforming two other improvements already made since Tuesday by 5e-8. At this pace, we will solve the Riemann Hypothesis by 2109! 😅
2
5
2
48
7,115
As I have noticed that other people have found similar small improvements, I created this Website to track Lean verified progress on the constant: gtimg.github.io/QuasiRiemann… Everyone is invited to submit a Lean proof for a better constant, let's see how low we can get this!
2
1
14
1,672
The first person to reach constant 1/2 is hereby cordially invited to a hot or cold beverage of their choice in Zurich. :D
1
13
954
GPT-6.1 Sol is the new #1 on MathArena! Cheaper and better than Astra.
Image from Jasper Dekoninck (@j_dekoninck)
21
52
3
517
27,925
We just added GPT-6 Sol to MathArena! It comes in only just behind GPT-6 Astra, at slightly lower cost. Despite its significantly lower cost per token, it reasons more, bringing its cost closer to Astra.
Image from Jasper Dekoninck (@j_dekoninck)
3
9
80
3,244
We just added DeepSeek-v4.1-Flash on MathArena! Not quite on the level of Qwen-3.8, but boy is it cheap.
Image from Jasper Dekoninck (@j_dekoninck)
8
4
3
117
18,340
We just added Fable 5.1 (low) and Astra (low) to MathArena! At low reasoning, both still have an impressive performance, but Astra remains more efficient and better.
Image from Jasper Dekoninck (@j_dekoninck)
3
6
77
3,245
Load more