INDEX

Explanations

common functional words that indicate various parts of speech or grammatical constructs

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

phrases starting with certain 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

 Herkunft

-0.26

 Einzelnen

-0.25

wiegend

-0.25

 Verfassung

-0.24

 Füßen

-0.24

 Betrach

-0.24

 Füße

-0.24

Origem

-0.24

 eigenes

-0.24

 falschen

-0.24

POSITIVE LOGITS

 otomatig

0.85

awtextra

0.84

 AssemblyCulture

0.80

 հղումներ

0.79

OGND

0.78

 kasarigan

0.75

IntoConstraints

0.73

 Italijani

0.71

 Administrativna

0.71

parsedMessage

0.71

Activations Density 0.010%

common functional words that indicate various parts of speech or grammatical constructs

phrases starting with certain words

No Comments

No Known Activations