INDEX

Explanations

phrases and actions related to printing and editing documents

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

use to create descriptions

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

 Select

-0.33

 three

-0.32

Les

-0.30

-0.29

Mou

-0.29

dot

-0.29

Sac

-0.28

sac

-0.28

lo

-0.28

ø

-0.28

POSITIVE LOGITS

 autorytatywna

0.84

verwijspagina

0.77

tonode

0.72

KommentareTeilen

0.70

 Chwiliwch

0.69

 Италијани

0.67

Tikang

0.66

webElementGuid

0.66

 հղումներ

0.63

 akár

0.62

Activations Density 0.058%

phrases and actions related to printing and editing documents

use to create descriptions

No Comments

No Known Activations