Skip to main content
MANIFOLD
Will the SotA open-source AI model in July 2026 use Test-Time Training?
11
Ṁ1kṀ985
Jul 31
35%
chance
Market context
AI
Technology
Technical AI Timelines
OpenAI
AI Impacts
Get
Ṁ1,000
to start trading!

People are also trading

April 2026 AI model releases
Will OpenAI officially launch any new publicly named AI model before May 1, 2026?
65% chance
If OpenAI open-sources o3-mini*, will it open-source an even more powerful model before July 2026?
66% chance
Will OpenAI announce a new GPT-5-level model before 1 July 2026?
97% chance
Will a publicly known AI model achieve an 80% time horizon that is an 1 hour and 30 minutes by September 2026?
90% chance
Will a publicly known AI model achieve an 80% time horizon of 3 weeks by April 2027?
28% chance
Will OpenAI release another open source LLM before end of 2026?
70% chance
Will open-source AI remain at least one year behind proprietary AI? (ACX, AI 2027 #4)
75% chance
AI model training time decreases fourfold by mid-2027?
36% chance
Will open-source AI win? (through 2028)
32% chance
Sort by:

Interestingly, SotA Lean provers use TTT, specifically AxiomMath, Kimina, and AlphaProof

I like this question

I think the definition of SotA will matter here. Since you can probably max out many of the benchmarks with normal training and some kind of memory module implemented with weight updates during testing may just be a bottleneck.

opened a Ṁ5,000 NO at 50% order

People are also trading

April 2026 AI model releases
Will OpenAI officially launch any new publicly named AI model before May 1, 2026?
65% chance
If OpenAI open-sources o3-mini*, will it open-source an even more powerful model before July 2026?
66% chance
Will OpenAI announce a new GPT-5-level model before 1 July 2026?
97% chance
Will a publicly known AI model achieve an 80% time horizon that is an 1 hour and 30 minutes by September 2026?
90% chance
Will a publicly known AI model achieve an 80% time horizon of 3 weeks by April 2027?
28% chance
Will OpenAI release another open source LLM before end of 2026?
70% chance
Will open-source AI remain at least one year behind proprietary AI? (ACX, AI 2027 #4)
75% chance
AI model training time decreases fourfold by mid-2027?
36% chance
Will open-source AI win? (through 2028)
32% chance