@Technocrat this is a great question, nice liquidity. however it would be nice to clarify for @traders on the resolution criteria and nature of the question to better facilitate trading. the question has a lot of potential that i feel currently isn’t being exercised.

for example, there are many forms of ELO. in the most common systems, a 100 elo increase refers to winning 64% of the points against an opponent 100 elo lower. however, this concept is much more loosely defined for super gms and chess engines, which possess less variability

and please clarify what “LLM agent” refers to. leading engine developers could and definitely use such agents. does human development with llm assistance or llm development with human assistance count? or does it have to be fully autonomous?

additionally, engine elo improvements are incremental. what if an llm improves performance by 60 elo and later another llm improves performance by another 60 elo?

just some thoughts.