INDEX

Explanations

names or proper nouns

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

names followed by specific suffixes

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



-0.75

 OMITBAD

-0.72

ագրություններ

-0.64

EDEFAULT

-0.61

+#+

-0.60

CloseOperation

-0.59

transQ

-0.59

Bakgrunnsstoff

-0.58

KommentareTeilen

-0.57

adaptiveStyles

-0.57

POSITIVE LOGITS

0.41

The

0.37

Position

0.36

Preconditions

0.35

is

0.35

하지

0.35

Flu

0.34

est

0.33

snprintf

0.33

Activations Density 0.534%

names or proper nouns

names followed by specific suffixes

No Comments

No Known Activations