INDEX

Explanations

punctuation and formatting characters in text

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

sending commands in code

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

évaluateur

-0.62

 BoxFit

-0.62

Tikang

-0.59

出版年

-0.59

 незавершена

-0.56

fromnode

-0.55

GOTREF

-0.54

Hentet

-0.54

AccessorTable

-0.54

 Chwiliwch

-0.54

POSITIVE LOGITS

runApp

0.38

 send

0.35

ggiare

0.33

 Италијани

0.33

 sent

0.31

gsSP

0.30

 interview

0.30

exitRule

0.30

urlencoded

0.29

Chham

0.29

Activations Density 0.005%

punctuation and formatting characters in text

sending commands in code

No Comments

No Known Activations