INDEX

Explanations

the first-person perspective in statements

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

'I' or 'My' followed by action or state

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

basicConfig

-0.49

RenderAtEndOf

-0.46

PostInfinity

-0.46

IsContent

-0.45

bitat

-0.44

 CRIME

-0.44

それだけ

-0.43

GenerationType

-0.43

 actionTypes

-0.43

 nonUne

-0.43

POSITIVE LOGITS

GEBURTSDATUM

0.48

ंदीखरीदारी

0.41

addContainerGap

0.40

mazoo

0.38

ícil

0.36

:✨

0.36

ientras

0.35



0.35

 الحره

0.34

ulongan

0.33

Activations Density 0.014%

the first-person perspective in statements

'I' or 'My' followed by action or state

No Comments

No Known Activations