INDEX

Explanations

phrases related to hyperlinks or clickable prompts

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

Билгалдахарш

-0.48

 sèche

-0.45

DataInputStream

-0.45

 refroidissement

-0.44

 trocken

-0.41

 mystère

-0.41

 commerciales

-0.40

 ladite

-0.39

 bedrag

-0.39

 débit

-0.38

POSITIVE LOGITS

 يتيمه

0.54

jstor

0.50

Diweddarwch

0.49

 препратки

0.49

 transfieras

0.47

LEncoder

0.47

tagHelperRunner

0.47

 ModelExpression

0.45

bootstrapcdn

0.45

ConverterFactory

0.44

Activations Density 0.173%

phrases related to hyperlinks or clickable prompts

No Comments

No Known Activations