INDEX

Explanations

frequently recurring phrases related to measurements or levels in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 doInBackground

-0.64

 Dentist

-0.47

 Charities

-0.46

 plástico

-0.45

textBox

-0.45

 cotton

-0.45

 abbraccio

-0.45

 rubber

-0.44

 recientes

-0.44

 Baumwolle

-0.44

POSITIVE LOGITS

 Level

1.31

 level

1.30

Level

1.29

 LEVEL

1.20

level

1.17

LEVEL

1.16

getLevel

1.14

 levels

1.11

 Levels

1.11

Levels

1.09

Activations Density 0.111%

frequently recurring phrases related to measurements or levels in various contexts

No Comments

No Known Activations