INDEX

Explanations

phrases indicating frequency or quantity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

hipster

-0.52

MessageOf

-0.50

addCriterion

-0.50

RTDA

-0.49

CppMethod

-0.49

Kit

-0.48

 uLocal

-0.47

oneofs

-0.47

 wireType

-0.47

Kit

-0.46

POSITIVE LOGITS

 swoich

0.40

 szüks

0.39

God

0.38

 pravi

0.38

 förs

0.37

zunehmen

0.36

 swoje

0.36

 resep

0.36

 dziecka

0.36

 saken

0.36

Activations Density 0.087%

phrases indicating frequency or quantity

No Comments

No Known Activations