← All parts of this equation

Equation 3 · Part 4 · How to Actually Compare Frontier AI Models Without Building a Misleading Leaderboard

Symbol H

s=g(θ,H,T,k,P,t)+ε,s = g(\theta, \mathcal{H}, T, k, \mathcal{P}, t) + \varepsilon,
H\mathcal{H}

What this part means

the harness.

Its job in the formula

H is one of the signed contributions combined to compute the quantity on the left.

Where the article explains it

a function of the weights θ\theta , the harness H\mathcal{H} , the sampling temperature T , the number of samples k and how they are aggregated, the prompt template P\mathcal{P} , the date t a snapshot was queried, plus sampling noise ε\varepsilon .

The passage around this formula

Formally, an observed score s is not a property of a model θ\theta alone. It is closer to s=g(θ,H,T,k,P,t)+εs = g(\theta, \mathcal{H}, T, k, \mathcal{P}, t) + \varepsilon. a function of the weights θ\theta , the harness H\mathcal{H} , the sampling temperature T , the number of samples k and how they are aggregated, the prompt template P\mathcal{P} , the date t a snapshot was queried, plus sampling noise ε\varepsilon . A citation that fixes θ\theta and leaves the other five free has not specified a comparable quantity.…

Read this part in the article →

Learn the underlying idea

A variable is a named place for a value. Its letter is a local label: x can mean position in one formula and a data point in another.

Open the illustrated variables: a letter stands for a value guide →

See this notation across published equations →

Sources cited in the surrounding passage

These citations provide research context; check each source for the exact claim it supports.