INDEX

Explanations

phrases indicating contrasts or comparisons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 ujednoznacz

-0.48

OOTDTY

-0.46

EnableWeb

-0.45

 biotic

-0.42

 uſed

-0.40

brigens

-0.40

 patin

-0.37

 Aplica

-0.37

 stim

-0.36

setof

-0.36

POSITIVE LOGITS

 merely

0.54

 mere

0.53

 лишь

0.44

ひたすら

0.43

 lediglich

0.43

 Mere

0.42

ofern

0.41

 vague

0.40

Mere

0.40

 barely

0.39

Activations Density 0.724%

phrases indicating contrasts or comparisons

No Comments

No Known Activations