INDEX

Explanations

proper nouns, particularly names of places and schools

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

ukone

-0.48

boutin

-0.46

principalColumn

-0.44

graphed

-0.43

 Obrador

-0.43

 NavController

-0.42

 muscular

-0.40

muscular

-0.38

AutoScaleMode

-0.38

よかった

-0.38

POSITIVE LOGITS

 تعدى

0.48

 disambiguazione

0.47

 lenker

0.46

 Italijanski

0.44

 VIAF

0.43

 invokingState

0.40

Pozri

0.40

VIAF

0.39

JsonKey

0.39

RTLR

0.39

Activations Density 0.199%

proper nouns, particularly names of places and schools

No Comments

No Known Activations