INDEX

Explanations

references to time-related phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

schirm

-0.46

ggak

-0.45

 Usaha

-0.44

 znam

-0.43

%)$

-0.43

impresa

-0.42

Connecticut

-0.42

Muhammad

-0.41

 Emmanuel

-0.41

 itſelf

-0.41

POSITIVE LOGITS

later

1.02

Later

0.95

 Later

0.93

 LATER

0.92

 later

0.90

 später

0.73

 späteren

0.70

earlier

0.63

+#+#

0.62

 فريبيس

0.62

Activations Density 0.055%

references to time-related phrases

No Comments

No Known Activations