INDEX

Explanations

elements related to recommendations or significant experiences people shouldn't miss

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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



-0.81

IsContent

-0.59

Tembelea

-0.51

 Meksiku

-0.49

XMLSchema

-0.48

dafx

-0.46

 Administrativna

-0.44

 estekak

-0.44

 Савезне

-0.44

setArguments

-0.44

POSITIVE LOGITS

OrBuilder

0.38

省

0.37

 Moser

0.36



0.35

able

0.35

 worthwhile

0.35

kable

0.35

σθαι

0.34

 الاطلاع

0.34

Klik

0.34

Activations Density 0.863%

elements related to recommendations or significant experiences people shouldn't miss

No Comments

No Known Activations