INDEX

Explanations

references to left and right orientations or directions

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

"left" and subsequent tokens

np_acts-logits-general · gemini-2.0-flash

left and right

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.30.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

 باد

-0.31

ッコ

-0.30

Million

-0.30

 وق

-0.30

 Million

-0.29

 Millennium

-0.29

hyperref

-0.28

 technic

-0.27

عت

-0.27

 MILLION

-0.27

POSITIVE LOGITS

AndEndTag

0.90

setLeft

0.77

 Left

0.75

Left

0.68

 AssemblyVersion

0.66

 sided

0.65

 noDo

0.63

Portail

0.63

EDEFAULT

0.63

 hand

0.61

Activations Density 0.187%

references to left and right orientations or directions

"left" and subsequent tokens

left and right

No Comments

No Known Activations