INDEX

Explanations

key elements of interactive dialogues and content structures

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

code or conversational marker `user`

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

delwed

-0.54

WithMany

-0.51

MessageTagHelper

-0.51

󠁴

-0.51

AndEndTag

-0.50

лтемелер

-0.48

الدراسه

-0.48

OGND

-0.47

 sumpay

-0.46

stylers

-0.46

POSITIVE LOGITS

0.40

produkte

0.39

][-

0.39

[]:

0.36

trice

0.36

awtextra

0.35

 ويكيپيديا

0.35

 condominium

0.35

 sulf

0.34

==(

0.34

Activations Density 0.001%

key elements of interactive dialogues and content structures

code or conversational marker `user`

No Comments

No Known Activations