INDEX

Explanations

technical terms and details related to specific processes or phenomena

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

step, condition, stored, stolen

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ंदीखरीदारी

-0.53

AddTagHelper

-0.52

webElementGuid

-0.50

 propOrder

-0.49

 Signalez

-0.49

 CanadaChoose

-0.49

oredCriteria

-0.48

MessageTagHelper

-0.47

FormTagHelper

-0.47

HtmlAttribute

-0.46

POSITIVE LOGITS

⟹

0.41

 للاسماء

0.40

漫

0.38

 IconData

0.37

 enti

0.37

?»

0.36

__':

0.36

梢

0.35

lebo

0.35

"/",

0.35

Activations Density 0.385%

technical terms and details related to specific processes or phenomena

step, condition, stored, stolen

No Comments

No Known Activations