INDEX

Explanations

references to knowledge acquisition or informational content

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

 ſte

-0.60

WillAppear

-0.59

 purpoſe

-0.58

GEBURTS

-0.57

elemField

-0.56

adpleegd

-0.56

CrossRef

-0.54

GIVEREF

-0.53

 ſind

-0.52

 ſta

-0.50

POSITIVE LOGITS

 knowledge

0.69

knowledge

0.56

'@

0.54

dataset

0.52

Knowledge

0.49

 Knowledge

0.48

 planning

0.46

"@

0.45

 Wissen

0.44

知识

0.43

Activations Density 0.083%

references to knowledge acquisition or informational content

No Comments

No Known Activations