INDEX

Explanations

calls to action, particularly phrases inviting the reader to click for more information

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

rungsseite

-0.72

tagHelperRunner

-0.60

 Audiodateien

-0.59

Erreferentziak

-0.57

 CURLOPT

-0.57

 linkovi

-0.57

 surla

-0.56

sidemargin

-0.56

Sklici

-0.55

StoreMessageInfo

-0.54

POSITIVE LOGITS

ſelf

0.51

 Conſ

0.49

 Chriftian

0.44

 ſche

0.42

 myſelf

0.41

 anſ

0.41

 Perſ

0.40

 viſ

0.40

 leaſt

0.40

 caisse

0.40

Activations Density 0.317%

calls to action, particularly phrases inviting the reader to click for more information

No Comments

No Known Activations