INDEX

Explanations

words that indicate certainty or speculation about a topic

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/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.27.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

 internetowa

-0.58

 nakalista

-0.53

المكان

-0.46

Abitanti

-0.46

RunWith

-0.45

WebControls

-0.45

 Entdecken

-0.44

liptic

-0.44

 otomatig

-0.43

✨:

-0.43

POSITIVE LOGITS

Personendaten

0.54

AxisAlignment

0.49

featureID

0.47

 esternos

0.45

 myſelf

0.43

 uſed

0.42

 purpoſe

0.40

 fubject

0.40

POM

0.40

POM

0.40

Activations Density 0.600%

words that indicate certainty or speculation about a topic

No Comments

No Known Activations