INDEX

Explanations

comparative terms indicating superiority or enhancements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 contextLoads

-0.56

 gear

-0.51

Mag

-0.50

BASELINE

-0.49

ContentAsync

-0.49

new

-0.49

 Haber

-0.47

auta

-0.46

Min

-0.45

 draft

-0.45

POSITIVE LOGITS

PreferredItem

0.68

 perfetta

0.62

ArrowToggle

0.61

 segreto

0.61

 Moderato

0.59

 disambiguazione

0.58

buttonBar

0.58

uxxxx

0.58

OGND

0.58

jooq

0.57

Activations Density 0.334%

comparative terms indicating superiority or enhancements

No Comments

No Known Activations