INDEX

Explanations

references to scientific journals or publications

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 Langford

-0.62

Sitz

-0.60

<()>

-0.60

 Lanka

-0.59

stddef

-0.59

xsi

-0.59

nito

-0.57

씨

-0.57

 Picasso

-0.56

 Winaray

-0.55

POSITIVE LOGITS

REV

1.53

Rev

1.42

REV

1.39

Rev

1.38

rev

1.27

rev

1.19

 Revisions

1.18

 Revolution

1.17

PhysRev

1.12

Revolution

1.10

Activations Density 0.084%

references to scientific journals or publications

No Comments

No Known Activations