Will a model costing >$30M be intentionally trained to be more mechanistically interpretable by end of 2027? (see desc)
Plus
30
Ṁ31362028
57%
chance
1D
1W
1M
ALL
Will any language model costing >$30M be intentionally trained with modifications to be more mechanistically interpretable by 2027?
If the training-scheme with modifications is more than a small amount better for capabilities than the [possibly next] capabilities-best training scheme at the time (available to the model trainers), it doesn't count for the purposes of this market.
Modifications to architecture, optimizer, etc. The modifications must be made with the intention to make the model more interpretable.
Costing >$30M is in my estimation, the cost does not need to be published.
I will not trade in this market.
This question is managed and resolved by Manifold.
Get
1,000and
3.00
Related questions
Related questions
AI: Will someone train a $10T model by 2100?
59% chance
Will mechanistic interpretability have more academic impact than representation engineering by the end of 2025?
72% chance
Will models be able to do the work of an AI researcher/engineer before 2027?
31% chance
Will it cost $30 to train a GPT-3 level model in 2030?
19% chance
Will interpretability be commonplace in physics papers relying on machine learning by the end of 2025?
10% chance
AI: Will someone train a $100M model by 2025?
85% chance
Will a GPT-3 quality model be trained for under $10.000 by 2030?
83% chance
Will a GPT-4 quality model be trained for under $10.000 by 2030?
78% chance
AI: Will someone train a $1B model by 2028?
81% chance
Will any 10 trillion+ parameter language model that follows instructions be released to the public before 2026?
38% chance