INDEX

Explanations

phrases related to professional etiquette and presentation

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

new section title

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

OGND

-1.03

AndEndTag

-1.01

 Normdatei

-0.93

 arşivlendi

-0.83

UserScript

-0.82

<unused14>

-0.81

 zuſammen

-0.81

[@BOS@]

-0.81

<unused41>

-0.81

<pad>

-0.80

POSITIVE LOGITS

 évêque

0.25

 allgemein

0.25

 catto

0.23

 insan

0.22

 zwar

0.21

userDao

0.20

Por

0.20

 dogs

0.20

Dir

0.20

 normalmente

0.20

Activations Density 0.067%

phrases related to professional etiquette and presentation

new section title

No Comments

No Known Activations