INDEX

Explanations

ellipsis and truncated thoughts or phrases

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

@[+][

-0.46

 koş

-0.46

SourceChecksum

-0.44

 alugar

-0.44

 kaldı

-0.43

 zasady

-0.42

 behaupten

-0.41

 Audiodateien

-0.41

 colgar

-0.41

 ayı

-0.40

POSITIVE LOGITS

webElementXpaths

0.50

sizeCache

0.46



0.44

IUrlHelper

0.43

ContentAsync

0.43

Портали

0.42

WithContext

0.41

oneofs

0.41

 ostavi

0.41

 IBOutlet

0.41

Activations Density 0.168%

ellipsis and truncated thoughts or phrases

No Comments

No Known Activations