INDEX

Explanations

phrases related to truth, change, and moral reasoning

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

AddTagHelper

-0.71

DeleteBehavior

-0.65

новништво

-0.64

 MainAxisSize

-0.63

BufferException

-0.62

Diweddarwch

-0.61

fjspx

-0.60

AndEndTag

-0.60

Personensuche

-0.59

 tartalomajánló

-0.59

POSITIVE LOGITS

 magically

0.36

ftagPool

0.35

ped

0.33

force

0.30

ste

0.30

 sfor

0.30

vai

0.30

ste

0.30

 Force

0.29

 trying

0.29

Activations Density 0.104%

phrases related to truth, change, and moral reasoning

No Comments

No Known Activations