INDEX

Explanations

proper nouns, particularly names of individuals

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

names of people

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

lgari

-0.44

材

-0.42

Personendaten

-0.42

 Milner

-0.41

zkod

-0.39

 kaarangay

-0.38

 both

-0.38

 Anlass

-0.37

ādā

-0.36

īdz

-0.36

POSITIVE LOGITS

 ExecuteAsync

0.65

tagHelperRunner

0.56

addContainerGap

0.53

WriteTagHelper

0.53

ValueStyle

0.47

 ब्रेकडाउन

0.47

__;

0.43

SBATCH

0.41

randomUUID

0.40

 تضيفلها

0.39

Activations Density 0.001%

proper nouns, particularly names of individuals

names of people

No Comments

No Known Activations