INDEX

Explanations

technical terms and components related to programming and software frameworks

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

featureID

-0.66

rawDesc

-0.66

Personendaten

-0.65

 rhestr

-0.59

resave

-0.57

transQ

-0.55

Diweddarwch

-0.54

ThroughAttribute

-0.54

tanleria

-0.52

 protoimpl

-0.52

POSITIVE LOGITS

 Conſ

0.47

 myſelf

0.45

IAm

0.44

 AppColors

0.43

 themſelves

0.43

 quæ

0.43

 Politique

0.43

StringCopy

0.42

ſelf

0.42

InThe

0.42

Activations Density 0.015%

technical terms and components related to programming and software frameworks

No Comments

No Known Activations