INDEX

Explanations

HTML tags or markup elements in the text

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

closing tag characters

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

featureID

-0.71

 typelib

-0.62

imachinery

-0.53

 nakalista

-0.53

fjspx

-0.49

شهاد

-0.48

 laikā

-0.48

 تانيه

-0.47

 wikipagina

-0.46

ServletException

-0.46

POSITIVE LOGITS

AutoScaleMode

0.43

<_>

0.40

InputTagHelper

0.39

MatInputModule

0.38

]")]

0.38

astic

0.36

 UserDao

0.36

DDS

0.35

entity

0.35

 entity

0.34

Activations Density 0.042%

HTML tags or markup elements in the text

closing tag characters

No Comments

No Known Activations