INDEX

Explanations

references to active components and their effectiveness 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 @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/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.0.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

Spécification

-0.52

 Bewußt

-0.51

Conclusão

-0.49

antaranya

-0.47

 Wahrheit

-0.46

Билгалдахарш

-0.46

 Tatsache

-0.46

 onCreateView

-0.45

 Grüsse

-0.44

 Könige

-0.43

POSITIVE LOGITS

 active

1.93

active

1.54

 Active

1.48

 ACTIVE

1.45

Active

1.40

ACTIVE

1.21

 актив

0.97

 aktive

0.95

 inactive

0.94

getActive

0.92

Activations Density 0.077%

references to active components and their effectiveness in various contexts

No Comments

No Known Activations