INDEX

Explanations

instances of the word "blow" and its variations, indicating references to negative impacts or failures

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Vezi

-0.63

 Alma

-0.56

 Middles

-0.55

geo

-0.55

 Solid

-0.54

 bronco

-0.54

 Дата

-0.53

Filmo

-0.53

flüs

-0.53

vábbi

-0.52

POSITIVE LOGITS

 blow

1.50

 blows

1.48

blow

1.31

 Blow

1.19

 blew

1.16

Blow

1.13

 blowing

1.08

 blown

1.02

blowing

1.00

 blowout

0.88

Activations Density 0.050%

instances of the word "blow" and its variations, indicating references to negative impacts or failures

No Known Activations