What will be true of OpenAI’s open-weight model?
45
1.4kṀ12k
Dec 31
1.4%
Uses Beyer Teacher/Student distillation ala Gemma 3
27%
Gets 1400+ Elo on lmarena.ai
59%
better than o3-mini on FrontierMath (with tools)

Context:

Feel free to add your own answers. If an answer is unclear you are welcome to ask for clarification from both me and the person who submitted the answer.

  • Update 2025-08-05 (PST) (AI summary of creator comment): Regarding an answer that the model will be 120B:

    • The creator has indicated this will likely resolve based on the actual number of weights/parameters.

    • If a model is named something like 'gpt-oss-120B' but does not technically have 120 billion weights, this answer would likely resolve to NO.

AI
OpenAI
Technical AI Timelines
AI Impacts
LLMs
Get
Ṁ1,000
to start trading!

People are also trading

Sort by:

whether it's 120B or not is kinda cursed tbh. Like there isn't 120B weights so probably NO, but it's literally called gpt-oss-120B

@Bayesian oh shit you're right it's 117B 🤦

"reasoning model" will be judged with reference to OpenAI documentation.

bought Ṁ10 YES

What’s the difference between image/language and vision/language?

bought Ṁ10 NO

@KJW_01294 I think image/language generates images whereas vision/language only takes them as input

I've chosen to bet on the market that references me, but assert that I will be truthful in my evaluation of it. To help others, I would have said no to Gemma 2/3, yes to PaliGemma, yes to DeepSeek-MoE, but no to DeepSeek V3 and DeepSeek R1

bought Ṁ10 YES

"parameters" means total parameters?

bought Ṁ30 NO

@JoshYou correct

People are also trading

© Manifold Markets, Inc.•Terms•Privacy