INDEX

Explanations

elements related to roles and responsibilities in a structured setting

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 głowy

-0.28

bł

-0.27

 scurt

-0.26

 możemy

-0.24

 жк

-0.24

 supérieures

-0.22

 verhält

-0.22

ValueStyle

-0.21

 الأم

-0.21

 event

-0.20

POSITIVE LOGITS

 betweenstory

0.82

 ویکی‌پدی

0.77

GraphicsUnit

0.74

IsMutable

0.73

Jîn

0.64

 platoon

0.64

iſchen

0.64

ﭜ

0.63

<unused52>

0.63

<unused8>

0.63

Activations Density 0.499%

elements related to roles and responsibilities in a structured setting

No Comments

No Known Activations