INDEX

Explanations

repeated mentions of the word "words."

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

 للاسماء

-0.59

 MainAxisSize

-0.58

Rptr

-0.56

tvguidetime

-0.53

*/].

-0.52

 poffe

-0.52

 prominently

-0.51

 mascot

-0.50

fjspx

-0.50

 multicultural

-0.50

POSITIVE LOGITS

 words

1.31

 Words

0.96

 palabras

0.79

Words

0.79

 woorden

0.77

 palavras

0.75

 słowa

0.73

 слова

0.70

 Worte

0.69

 Worten

0.68

Activations Density 0.140%

repeated mentions of the word "words."

No Comments

No Known Activations