← All parts of this equation

Equation 1 · Part 5 · Adapters, Native Pretraining, and Unified Tokens: The Main Multimodal Architectures, Compared

Symbol theta_bridge

θ(t+1)=θ(t)−η⋅1 ⁣[θ∈θbridge]⊙∇θ L(θ(t))\theta^{(t+1)} = \theta^{(t)} - \eta \cdot \mathbb{1}\!\left[\theta \in \theta_{\text{bridge}}\right] \odot \nabla_\theta \, \mathcal{L}(\theta^{(t)})
θbridge\theta_{\text{bridge}}

What this part means

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

Its job in the formula

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

The passage around this formula

…and everything the language side ever learns about the image has to fit through it. This is the formal shape of “freeze the backbones”: the gradient update is masked to zero everywhere except the bridge parameters θbridge\theta_{\text{bridge}} , so |θbridge\theta_{\text{bridge}}| — a few hundred million parameters in BLIP-2’s case, dramatically fewer than either tower — is the entire quantity being optimised, and θvision\theta_{\text{vision}} and θLM\theta_{\text{LM}} contribute a forward pass but never a gradient.

Read this part in the article →

Learn the underlying idea

A subscript is a label attached below a symbol. It often selects a time step, component, category, or member of a sequence.

Open the illustrated subscripts: which member of a family? 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.