INDEX

Explanations

references to wolves

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

=&

-0.42

CHD

-0.41

DataAnnotations

-0.39

('&

-0.39

',

-0.38

чери

-0.38

/**

-0.38

CCR

-0.38

=")

-0.37

aysia

-0.36

POSITIVE LOGITS

Wolf

1.27

 Wolf

1.26

 WOLF

1.16

 wolf

1.13

wolf

1.12

Wolves

1.03

 Wolves

1.03

 Wolfs

1.02

 wolves

1.01

wolves

0.94

Activations Density 0.004%

references to wolves

No Comments

No Known Activations