INDEX

Explanations

special characters or symbols used for formatting or denoting specific elements in text

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

user turn separators

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

DockStyle

-0.57

новниш

-0.54

RenderAtEndOf

-0.43

圣节

-0.42

exitRule

-0.41

RectangleBorder

-0.40

 aufzu

-0.40

RUnlock

-0.40

CPtr

-0.40

AntiForgeryToken

-0.40

POSITIVE LOGITS

 informée

0.60

 للمعارف

0.53

 ***!

0.47

 autorytatywna

0.42

:✨

0.42

 Wesen

0.42

 förra

0.41

rant

0.41

VYMaps

0.41

Portale

0.41

Activations Density 0.000%

special characters or symbols used for formatting or denoting specific elements in text

user turn separators

No Comments

No Known Activations