INDEX

Explanations

significant keywords related to proposals and comparisons

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

different followed by specific noun

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

ۣ

-0.46

 initComponents

-0.39

 Rising

-0.36

 tunt

-0.35

 kanya

-0.33

 rises

-0.33

 storm

-0.32

 setTotal

-0.32

dienste

-0.32

 stabilisation

-0.32

POSITIVE LOGITS

FunctionFlags

0.59

 Italijanski

0.59

 lenker

0.58

 betweenstory

0.58

Населення

0.56

 transférez

0.55

 Италијани

0.54

лтамалар

0.54

 beginnetje

0.53

invokeLater

0.53

Activations Density 0.103%

significant keywords related to proposals and comparisons

different followed by specific noun

No Comments

No Known Activations