INDEX

Explanations

prominent elements of writing, such as titles or leading phrases that introduce concepts or questions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 resourceCulture

-0.77

AndEndTag

-0.67

 المعيارى

-0.64

DeleteBehavior

-0.63

ValueStyle

-0.63

 بيها

-0.62

AccessorTable

-0.60

DoubleQuotes

-0.60

 تضيفلها

-0.60

IUrlHelper

-0.60

POSITIVE LOGITS

lámpara

0.35

 débar

0.32

 independencia

0.32

 Autorisations

0.32

protetor

0.31

 generación

0.31

 nonUne

0.30

 índice

0.28

@[+][

0.28

Activations Density 0.442%

prominent elements of writing, such as titles or leading phrases that introduce concepts or questions

No Comments

No Known Activations