INDEX

Explanations

unique identifiers associated with data entries

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

parsedMessage

-0.54

 يتيمه

-0.46

 TabLayout

-0.42

íslu

-0.42

alamualaikum

-0.42

ThroughAttribute

-0.41

Exped

-0.41

eligible

-0.40

scriptcase

-0.40

 تضيفلها

-0.40

POSITIVE LOGITS

 Unique

0.60

 uniqueness

0.57

 unique

0.57

Unique

0.54

unique

0.50

 UNIQUE

0.46

➟

0.45

 unieke

0.45

 уника

0.41

 uniques

0.41

Activations Density 0.610%

unique identifiers associated with data entries

No Comments

No Known Activations