INDEX

Explanations

references to programming or technical assistance

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

Personendaten

-0.59

MessageOf

-0.48

 дописавши

-0.47

ɵɵ

-0.46

Havolalar

-0.44

 محفوظة

-0.42

transQ

-0.42

 kaarangay

-0.42

tableFuture

-0.41

IntoConstraints

-0.41

POSITIVE LOGITS

 AssemblyTitle

0.66

helper

0.61

 المعيارى

0.51

 Helper

0.49

llac

0.48

ViewFeatures

0.48

PreferredItem

0.43

toMatchSnapshot

0.42

 helper

0.41

SharedDtor

0.40

Activations Density 0.000%

references to programming or technical assistance

No Comments

No Known Activations