INDEX

Explanations

language associated with legal and procedurally relevant terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

Cubit

-0.45

StoryboardSegue

-0.43

 okuyayım

-0.41

 couvre

-0.40

Opin

-0.40

IsPostBack

-0.40

adə

-0.40

 audits

-0.40

 planea

-0.39

 ease

-0.39

POSITIVE LOGITS

 oprot

0.50

awtextra

0.50

 Normdatei

0.49

 esternos

0.48

 IMMEDIATE

0.48

plish

0.47

 Immediate

0.46

 nonUne

0.46

 immediate

0.45

 instant

0.44

Activations Density 0.760%

language associated with legal and procedurally relevant terms

No Comments

No Known Activations