INDEX

Explanations

phrases related to writing and publishing projects

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

future actions or ongoing work

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

 uſed

-0.44

Especificación

-0.36

 incendios

-0.36

 amarillas

-0.36

 secado

-0.35

 increí

-0.34

 sanitarias

-0.33

 pères

-0.33

 confusión

-0.33

 étn

-0.32

POSITIVE LOGITS

StoryboardSegue

0.78

ronpa

0.68

 صوتيه

0.62

Искәрмәләр

0.59

StructEnd

0.56

Jereo

0.55

utilisons

0.52

 للاسماء

0.52

ResponseWriter

0.51

fillType

0.50

Activations Density 0.017%

phrases related to writing and publishing projects

future actions or ongoing work

No Comments

No Known Activations