INDEX

Explanations

references to heads and forms of headwear

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

AMAHA

-0.57

readystatechange

-0.56

 CommonModule

-0.55

engesch

-0.54

 WebDriverWait

-0.51

 Marit

-0.51

 corridors

-0.50

väll

-0.49

 McNally

-0.48

 gratification

-0.47

POSITIVE LOGITS

 head

1.43

 heads

1.36

 Head

1.30

Head

1.23

 HEAD

1.19

head

1.18

Heads

1.16

 Heads

1.15

heads

1.12

HEAD

1.10

Activations Density 0.956%

references to heads and forms of headwear

No Comments

No Known Activations