INDEX

Explanations

metadata related to images and videos

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

file types and identifiers

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

InjectMocks

-0.40

IsPostBack

-0.34

 scolaire

-0.32

HomeAsUpEnabled

-0.31

tidaknya

-0.31

 drept

-0.30

 quedado

-0.30

 empê

-0.30

ĝis

-0.30

Finance

-0.29

POSITIVE LOGITS

httphttps

0.71

ConstraintMaker

0.69

 فريبيس

0.66

complexContent

0.65

 esternos

0.63

SequentialGroup

0.62

 estekak

0.60

bootstrapcdn

0.60

 Signalez

0.57

󠁣

0.57

Activations Density 0.002%

metadata related to images and videos

file types and identifiers

No Comments

No Known Activations