INDEX

Explanations

identifiers, codes, or technical specifications related to programming or system errors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 distintos

-0.29

zem

-0.29

fos

-0.28

ത്ത

-0.27

bas

-0.26

ฑ

-0.26

 личности

-0.25

nay

-0.25

 personnes

-0.24

akaian

-0.24

POSITIVE LOGITS

SequentialGroup

0.58

ftagPool

0.57

niſſe

0.57

findpost

0.55

ſſung

0.54

ⵌ

0.54

colhead

0.53

 imagui

0.53

<unused14>

0.52

STELL

0.52

Activations Density 0.429%

identifiers, codes, or technical specifications related to programming or system errors

No Comments

No Known Activations