INDEX

Explanations

terms related to data handling, including expressions for combining and filtering data or objects

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

ی

-0.42

-0.41

an

-0.36

 than

-0.33

-0.32

be

-0.31

or

-0.30

for

-0.30

POSITIVE LOGITS

AddTagHelper

0.76

########.

0.72

LookAnd

0.71

 kasarigan

0.68

 betweenstory

0.66

gnition

0.63

MigrationBuilder

0.61

 Italijanski

0.61

 صوتيه

0.61

uxxxx

0.60

Activations Density 0.160%

terms related to data handling, including expressions for combining and filtering data or objects

No Comments

No Known Activations