INDEX

Explanations

references to layers or layered structures in various contexts

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.07

Î±Î¹

-0.06

Ð¾Ð´Ð¾

-0.06

 Halk

-0.06

ucht

-0.06

ãĥ³ãĤ¹

-0.06

igli

-0.06

Î¼Î¿

-0.06

æķ£

-0.06

POSITIVE LOGITS

 layers

0.23

 layer

0.20

 Layers

0.20

layers

0.18

Layers

0.17

 Layer

0.17

-layer

0.17

layer

0.17

_layers

0.16

Layer

0.16

Activations Density 0.057%