INDEX

Explanations

instances of ellipses, indicating pauses or omissions in speech or thought

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

 BoxDecoration

-0.58

 Infórmanos

-0.57

rsiniz

-0.55

AddTagHelper

-0.54

Datuak

-0.54

 Citadel

-0.52

各样的

-0.51

urtstag

-0.50

oflavin

-0.49

strix

-0.48

POSITIVE LOGITS

...

1.10

…

1.01

ldots

0.96

...

0.91

....

0.89

!..

0.84

…

0.84

?...

0.84

!...

0.84

….

0.82

Activations Density 0.162%

instances of ellipses, indicating pauses or omissions in speech or thought

No Comments

No Known Activations