INDEX

Explanations

phrases indicating actions and their results in a procedural context

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

ViewImports

-0.49

MessageTagHelper

-0.46

 ſch

-0.45



-0.44

 playmobil

-0.43

 PAGER

-0.41

bap

-0.41

IUrlHelper

-0.40

 gecko

-0.40

 فريبيس

-0.39

POSITIVE LOGITS

DispatchToProps

0.52

parseDouble

0.48

 Unterschiede

0.46

onViewCreated

0.45

AddTagHelper

0.45

endpush

0.43

 menim

0.42

 asupra

0.42

basicConfig

0.42

 CreateTagHelper

0.42

Activations Density 0.778%

phrases indicating actions and their results in a procedural context

No Comments

No Known Activations