Do you think Mixture of Expert (MoE) transformer models are generally more human interpretable than dense transformers? | Manifold

Do you think Mixture of Expert (MoE) transformer models are generally more human interpretable than dense transformers?

16

Never closes

Yes

No

Results

If you'd like to trade based on confidence instead of just YES/NO, here's a market:

Market context

Technical AI Safety

Mechanistic interpretability

Get

1,000

to start trading!

People are also trading

Will mechanistic/transformer interpretability [eg Neel Nanda] end up affecting p(doom) more than 5%?

Will the most capable, public multimodal model at the end of 2027 in my judgement use a transformer-like architecture?

Will the first AGI be a transformer model?

Related questions

Will mechanistic/transformer interpretability [eg Neel Nanda] end up affecting p(doom) more than 5%?

Will the most capable, public multimodal model at the end of 2027 in my judgement use a transformer-like architecture?

Will the first AGI be a transformer model?

© Manifold Markets, Inc.•Terms•Privacy