INDEX

Explanations

phrases that indicate insight or previews into experiences or settings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 berita

-0.48

 charge

-0.47

createQuery

-0.47

 need

-0.45

มาก

-0.45

mist

-0.44

Damit

-0.44

 morts

-0.44

シン

-0.43

 namanya

-0.42

POSITIVE LOGITS

 glimpse

1.25

 glimpses

1.25

 Einblick

1.16

ChildScrollView

1.07

 للمعارف

1.03

tagext

1.00

NOPQRST

0.99

SourceChecksum

0.98

DockStyle

0.98

 Peek

0.98

Activations Density 0.152%

phrases that indicate insight or previews into experiences or settings

No Comments

No Known Activations