INDEX

Explanations

code-related terminology and structures

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

GOTREF

-0.73

OGND

-0.68

 kasarigan

-0.65

Tikang

-0.65

 للمعارف

-0.60

 Paglinawan

-0.60

mybatisplus

-0.58

 AssemblyCulture

-0.57

GIVEREF

-0.56

SharedCtor

-0.56

POSITIVE LOGITS

putExtra

0.33

 enví

0.32

 ease

0.31

uib

0.30

 apologize

0.30

 tell

0.30

 Wall

0.30

layoutControl

0.30

]<<"

0.29

 relayed

0.29

Activations Density 0.485%

code-related terminology and structures

No Comments

No Known Activations