INDEX

Explanations

repeated phrases or references to 'line.'

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

line of

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

 Fabulous

-0.45

 poroz

-0.45

 desay

-0.43

 spectacular

-0.43

 Scha

-0.42

 Erst

-0.42

 Carnaval

-0.42

 highly

-0.41

 smallest

-0.41

 youngest

-0.41

POSITIVE LOGITS

Line

1.34

 Line

1.30

 LINE

1.24

 line

1.22

line

1.20

LINE

1.18

 Lines

1.03

Lines

1.02

 LINES

0.97

lines

0.96

Activations Density 0.047%

repeated phrases or references to 'line.'

line of

No Comments

No Known Activations