INDEX

Explanations

intensifiers or adverbs that amplify the degree of an adjective or adverb

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

very + adjective

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

uglify

-0.33

CHT

-0.32

 Staf

-0.30

 Supremo

-0.29

 suprême

-0.29

</tfoot>

-0.28

 nied

-0.28

ugc

-0.28

CWE

-0.28

 wali

-0.27

POSITIVE LOGITS

GEBURTSDATUM

0.65

 surla

0.65

 מאוד

0.59

 BoxDecoration

0.58

 LoginForm

0.58

Hentet

0.57

 kasarigan

0.57

<unused41>

0.56

tvguidetime

0.56

 EconPapers

0.55

Activations Density 0.023%

intensifiers or adverbs that amplify the degree of an adjective or adverb

very + adjective

No Comments

No Known Activations