INDEX

Explanations

expressions related to impactful speech or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.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

 ब्रेकडाउन

-0.46

Portály

-0.39

matchCondition

-0.39

 للمعارف

-0.38

łaś

-0.37

 שוליים

-0.37

 worin

-0.36

Portail

-0.35

untung

-0.34

 légiti

-0.33

POSITIVE LOGITS

 indelible

1.35

 impression

1.26

 imprint

1.24

 imprinted

1.23

 impressions

1.20

 lasting

1.18

impression

1.16

印象

1.15

 memorable

1.14

 unforgettable

1.13

Activations Density 0.266%

expressions related to impactful speech or statements

No Comments

No Known Activations