INDEX

Explanations

phrases that indicate viewing or referencing content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

MP

-0.35

mp

-0.34

MP

-0.32

 sett

-0.32

lder

-0.30

 fact

-0.30

dar

-0.30

Inc

-0.29

何と

-0.29

 Liberties

-0.28

POSITIVE LOGITS

 nahilalakip

0.74

ThroughAttribute

0.69

 snippetHide

0.68

 firsthand

0.67

featureID

0.66

setViewportView

0.66

OGND

0.65

rungsseite

0.65

Saw

0.65

saw

0.64

Activations Density 0.146%

phrases that indicate viewing or referencing content

No Comments

No Known Activations