INDEX

Explanations

instances of the word "notice" and its variations in a context related to observation or awareness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

ycznego

-0.49

Acceso

-0.47

 electrónico

-0.43

 electrico

-0.42

YMS

-0.42

ersatz

-0.41

ądź

-0.41

醐

-0.40

bitkan

-0.40

 něm

-0.40

POSITIVE LOGITS

 noticed

1.24

 noticing

1.20

 noticeable

1.07

noticed

1.07

 noted

1.05

 notice

0.96

 notices

0.92

noted

0.90

 observed

0.88

notice

0.88

Activations Density 0.341%

instances of the word "notice" and its variations in a context related to observation or awareness

No Comments

No Known Activations