INDEX

Explanations

terms related to user authentication and credentials

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

password and 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

IntoConstraints

-0.77

+#+#

-0.75

 propOrder

-0.69

 estekak

-0.68

#+#

-0.65

AndEndTag

-0.64

EDEFAULT

-0.64

richTextPanel

-0.63



-0.61

httphttps

-0.60

POSITIVE LOGITS

 especialidad

0.36

 mulut

0.31

 misión

0.30

 invitado

0.30

 enseñanza

0.30

 movilidad

0.29

 desaparecido

0.28

 vegetación

0.28

 antigüedad

0.28

/*

0.28

Activations Density 0.002%

terms related to user authentication and credentials

password and code

No Comments

No Known Activations