INDEX

Explanations

phrases indicating prior knowledge or confirmation of information

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

shafen

-0.35

itaire

-0.34

 Campaigns

-0.34

干

-0.33

干

-0.32

 apro

-0.31

appro

-0.31

BMI

-0.31

Circa

-0.30

Dry

-0.30

POSITIVE LOGITS

AsUp

0.59

 CURIAM

0.58

 Infórmanos

0.57

setVerticalGroup

0.54

 Administrativna

0.51

IndentedString

0.50

 already

0.50

webElementXpaths

0.50

tagext

0.49

VYMaps

0.48

Activations Density 0.462%

phrases indicating prior knowledge or confirmation of information

No Comments

No Known Activations