INDEX

Explanations

numerical data, particularly dates and statistics

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

years

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 ProtoMessage

-0.68

Tikang

-0.68

 transfieras

-0.67

 يتيمه

-0.65

expandindo

-0.60

FunctionFlags

-0.59

MigrationBuilder

-0.59

 resourceCulture

-0.56

OGND

-0.54

RotationOrder

-0.54

POSITIVE LOGITS

 strijd

0.36

ียง

0.34

IsPostBack

0.32

 almoço

0.31

DIPSETTING

0.30

 bebê

0.30

 célèbres

0.30

zbęd

0.29

 років

0.29

Atentamente

0.29

Activations Density 0.382%

numerical data, particularly dates and statistics

years

No Comments

No Known Activations