INDEX

Explanations

references to documentation or technical specifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

as

-0.40

pa

-0.39

 Lehmann

-0.37

ks

-0.35

return

-0.33

 Illuminate

-0.33

 hòa

-0.32

se

-0.32

ловек

-0.32

POSITIVE LOGITS

AndEndTag

0.75

 CreateTagHelper

0.69

MessageOf

0.68

 AssemblyTitle

0.65

Diweddarwch

0.64

parsedMessage

0.61

لينكات

0.60

 Taktlose

0.60

fromnode

0.60

IVEREF

0.60

Activations Density 0.000%

references to documentation or technical specifications

No Comments

No Known Activations