INDEX

Explanations

mathematical operations and expressions

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

end of the

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

crí

-0.28

 gegründet

-0.26

 localObject

-0.25

 удалить

-0.24

žití

-0.23

 yaptığı

-0.23

žit

-0.23

Entity

-0.23

 procesos

-0.22

 Luar

-0.22

POSITIVE LOGITS

 ModelExpression

0.86

GTCX

0.74

 للاسماء

0.73

AndEndTag

0.72

 EconPapers

0.70

findpost

0.68

zzleHttp

0.68

<unused8>

0.67

[@BOS@]

0.67

<pad>

0.67

Activations Density 0.132%

mathematical operations and expressions

end of the

No Comments

No Known Activations