INDEX

Explanations

phrases related to discussions, analyses, and research topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/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.23.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

 bezeichneter

-0.46

LEGGI

-0.44

rungsseite

-0.42

áctenos

-0.42

InputTagHelper

-0.41

Hentet

-0.41

 Rois

-0.41

 getItemCount

-0.40

conomie

-0.40

bitField

-0.40

POSITIVE LOGITS

 Numerade

0.47

Здра

0.40

 Felicia

0.40

 bang

0.39

yargs

0.39

endswith

0.38

GEBURTS

0.37

 firing

0.37

 ✭✭

0.36

ISO

0.36

Activations Density 0.288%

phrases related to discussions, analyses, and research topics

No Comments

No Known Activations