INDEX

Explanations

words that indicate relationships and connections in a contextual framework

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

expressions involving 'over' and mathematical terms

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 betweenstory

-0.64

+#+

-0.53

GEBURTSDATUM

-0.46

AddHtmlAttribute

-0.45

 Händen

-0.44

mäßig

-0.44

 numberOfRows

-0.44

 wireType

-0.43

/*++

-0.43

amoan

-0.42

POSITIVE LOGITS

enumi

0.48

0.42

ベン

0.39

minist

0.37

自分も

0.37

 Definition

0.36

常

0.36

 sanity

0.36

 Zeug

0.36

гаре

0.35

Activations Density 0.055%

words that indicate relationships and connections in a contextual framework

expressions involving 'over' and mathematical terms

No Comments

No Known Activations