INDEX

Explanations

expressions indicating degrees of certainty or levels of emphasis

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

common grammatical words

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

 AssemblyCulture

-0.78

GEBURTS

-0.78

iſen

-0.75

脚注の使い方

-0.73

ésultats

-0.71

itsubishi

-0.69

Tembelea

-0.69

ſehen

-0.69

ſchen

-0.68

ſcher

-0.68

POSITIVE LOGITS

or

0.41

and

0.33

in

0.32

for

0.31

at

0.31

 under

0.30

 during

0.29

if

0.28

 even

0.28

Activations Density 0.055%

expressions indicating degrees of certainty or levels of emphasis

common grammatical words

No Comments

No Known Activations