INDEX

Explanations

elements related to specific features and structures in data or formats

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 13-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_13/width_16k/average_l0_132

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.13.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

ThroughAttribute

-0.53

PerformLayout

-0.49

bitField

-0.45

⟬

-0.42

 estekak

-0.41

 Ayer

-0.39

Personendaten

-0.38

Hentet

-0.38

UnusedPrivate

-0.37

 snippetHide

-0.37

POSITIVE LOGITS

 onCreateView

0.48

 ModelExpression

0.46

 createState

0.43

EndContext

0.42

 članak

0.41

 PARTICIP

0.40

 Merri

0.40

ümmer

0.39

 Putih

0.39

Fuß

0.38

Activations Density 1.249%

elements related to specific features and structures in data or formats

No Comments

No Known Activations