INDEX

Explanations

geographic locations and place names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

 lenker

-0.71

出版年

-0.65

windowFixed

-0.65

 Wikimedijinoj

-0.60

Oregon

-0.57

Portland

-0.57

Seattle

-0.56

 OREGON

-0.54

 Oregon

-0.53

 Portland

-0.53

POSITIVE LOGITS

 Olympic

0.41

 Shelton

0.37

vdots

0.36

Parcelize

0.36

 Pend

0.35

OLY

0.35

 affid

0.34

Gig

0.33

 Olímp

0.33

 Quil

0.33

Activations Density 0.351%

geographic locations and place names

No Comments

No Known Activations