INDEX

Explanations

references to assumptions and clarifications regarding understanding and conditions

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

you are like your

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

/*

-0.52

Tembelea

-0.44

 dortmund

-0.39

 sarili

-0.39

icoli

-0.38

 maailman

-0.37

WriteTagHelper

-0.37

Kanpo

-0.36

цезда

-0.36

Portale

-0.36

POSITIVE LOGITS

transQ

0.59

enment

0.45

 oprot

0.43

Становништво

0.43

enumi

0.43

hoeddwyd

0.42

 estekak

0.42

 bang

0.42

]")]

0.41

SequentialGroup

0.39

Activations Density 0.059%

references to assumptions and clarifications regarding understanding and conditions

you are like your

No Comments

No Known Activations