INDEX

Explanations

key concepts and entities related to documentation and projects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

at

-0.38

of

-0.37

as

-0.36

↵

-0.30

-0.28

cientos

-0.27

er

-0.27

’

-0.27

<h2>

-0.27

qu

-0.27

POSITIVE LOGITS

 faſt

0.87

 ſte

0.81

 يتيمه

0.81

OGND

0.79

 queſta

0.78

 surla

0.76

 houſe

0.75

Personendaten

0.75

principalTable

0.75

 leſs

0.75

Activations Density 0.127%

key concepts and entities related to documentation and projects

No Comments

No Known Activations