INDEX

Explanations

phrases that express purpose or intention

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

ſelves

-0.65

ſelf

-0.62

ImageField

-0.58

 juſ

-0.55

StructEnd

-0.52

 pleaſure

-0.50

ToBounds

-0.49

 BorderSide

-0.49

 deſt

-0.48

principalColumn

-0.48

POSITIVE LOGITS

 requires

0.44

Síguenos

0.41

 firstly

0.40

詳しくは

0.39

GEBURTSDATUM

0.38

するには

0.38

Firstly

0.38

 first

0.37

plz

0.37

Requires

0.37

Activations Density 0.728%

phrases that express purpose or intention

No Comments

No Known Activations