INDEX

Explanations

proper nouns, particularly names and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

WebElementEntity

-0.53

Personensuche

-0.47

.*")]

-0.46

MOL

-0.42

WebVitals

-0.42

interopRequire

-0.42

queryInterface

-0.42

Portail

-0.42

/*

-0.40

🏻

-0.38

POSITIVE LOGITS

 Picchu

0.48

ftagPool

0.44

ch

0.42

basicConfig

0.42

roch

0.42

oches

0.41

URCH

0.41

Lach

0.41

illerato

0.41

angelog

0.41

Activations Density 0.164%

proper nouns, particularly names and related terms

No Comments

No Known Activations