INDEX

Explanations

the prefix "re-" indicating repetition or returning to a previous state

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

Personendaten

-0.63

 defaultstate

-0.57

AddTagHelper

-0.55

 betweenstory

-0.54

setVerticalGroup

-0.54

dafx

-0.52

matchCondition

-0.50

WithIOException

-0.50

CloseOperation

-0.49

ftagPool

-0.49

POSITIVE LOGITS

re

0.78

re

0.40

 jengibre

0.38

 emergency

0.37

[];

0.37

 spä

0.36

 अव

0.36

Legături

0.36

Przypisy

0.36

 upper

0.35

Activations Density 0.127%

the prefix "re-" indicating repetition or returning to a previous state

No Comments

No Known Activations