INDEX

Explanations

phrases and terms related to quality, convenience, and informality

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

describing qualities or categories

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

 виправивши

-0.32

enumi

-0.31

 censiti

-0.30

TV

-0.30

 onPostExecute

-0.29

Cyfeiriadau

-0.29

makeText

-0.29

omitempty

-0.27

RemindMe

-0.27

ilirsiniz

-0.26

POSITIVE LOGITS

 myſelf

0.51

новниш

0.50

hirt

0.50

 poffe

0.50

archar

0.50



0.50

 ***!

0.50

Spolja

0.49

 graded

0.49

Autoritní

0.49

Activations Density 0.069%

phrases and terms related to quality, convenience, and informality

describing qualities or categories

No Comments

No Known Activations