INDEX

Explanations

utilitarian expressions related to usage and value in various contexts

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

used to find

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

(&:

-0.30

GOTREF

-0.29

łaszcza

-0.27

tête

-0.24

ViewFeatures

-0.24

oacute

-0.24

 Acacia

-0.24

 Reflect

-0.23

 Another

-0.23

!("{}",

-0.23

POSITIVE LOGITS

RectangleBorder

0.63

Jeografia

0.57

providedIn

0.57

tagHelperRunner

0.57

mobileqq

0.55

 HasFactory

0.55



0.54

 müſſen

0.54

tonode

0.54

iſchen

0.53

Activations Density 0.015%

utilitarian expressions related to usage and value in various contexts

used to find

No Comments

No Known Activations