INDEX

Explanations

numerical values or mathematical expressions

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 gezet

-0.56

нгред

-0.54

 zoude

-0.51

 gezegd

-0.51

actéristique

-0.51

alugar

-0.51

>\<^

-0.50

 prioritise

-0.50

 Wallflower

-0.50

ustimmung

-0.50

POSITIVE LOGITS

0.81

awtextra

0.51

↵↵

0.51

 Taktlose

0.49

<h3>

0.48

<h2>

0.48

 ExecuteAsync

0.47

<strong>

0.47

TagMode

0.47

<h1>

0.47

Activations Density 0.253%

numerical values or mathematical expressions

No Comments

No Known Activations